A curated weekday guide to major news and developments over the weekend. Here’s today’s news:

IRAN WAR

The U.S. government has put employees in the Middle East on heightened alert and moved to reduce personnel levels at embassies and consulates because intelligence shows Iran is in the advanced stage of planning attacks against U.S. citizens, U.S. officials told the New York Times on Friday. The U.S. Director of National Intelligence Jay Clayton raised alarms to U.S. Secretary of State Marco Rubio in a memo obtained by Axios. Edward Wong, Jonathan Swan, and Julian E. Barnes report; Barak Ravid reports.

Iranian Ambassador to Pakistan Reza Amiri Moghadam on Saturday said Iran remains open to talks with the United States under a Pakistan-brokered interim deal reached in June to end the conflict. Amiri Moghadam said Iran recently conveyed through Pakistan that it was prepared to reopen the Strait of Hormuz if the United States lifted its naval blockade, and accused President Trump of rejecting the proposal by insisting the blockade would continue. Munir Ahmed reports for AP News.

IRAN WAR – YEMEN

Yemen’s Iran-aligned Houthis today said they had attacked Saudi Arabia’s East-West crude oil pumping stations and Riyadh airport. There was no immediate Saudi confirmation of the Houthi attacks. The Saudi-led coalition in Yemen earlier said it had destroyed two ballistic missile launchers in Yemen’s capital Sanaa, including one used to fire a missile towards Saudi’s capital Riyadh that the coalition said it had intercepted. Yomna Ehab and Eman Abouhassira report for Reuters.

ISRAEL-HAMAS WAR

The Israeli military overnight on Saturday began a crackdown in the West Bank city of Hebron, a day after Palestinian gunmen killed two Israeli soldiers in the city. Israeli Prime Minister Benjamin Netanyahu vowed to retaliate for the deadly shooting. The IDF were searching homes, making arrests, and using loudspeakers to tell residents of Hebron’s Deir Baha neighborhood to leave, according to Hebron’s mayor Yousef Jabari. Aaron Boxerman and Fatima AbdulKarim report for the New York Times.

WEST BANK DEVELOPMENTS

Italy’s Foreign Ministry on Saturday said five Italian activists who were detained by Israeli forces in the West Bank this week were released and repatriated. The women were detained along with other international activists on Thursday by Israeli police in Kiryat Arba while they sought to protect Palestinian farmers during the olive harvest in the Israel-occupied territory. Cristina Carlevaro reports for Reuters.

RUSSIA-UKRAINE WAR

U.S. First Lady Melania Trump on Friday said she brokered a “localized” ceasefire agreement between Russia and Ukraine in the Ukrainian city of Oleshky, slated to begin on Oct. 29. Melania said that “without my intervention, the possibility to bring humanitarian aid into the city and to evacuate the elderly would be impossible.” Gregory Svirnovskiy reports for POLITICO.

Russia’s Defense Ministry today accused Ukraine of repeatedly striking the Russian-controlled town of Oleshky in southern Ukraine ahead of a planned local ceasefire this week. The ministry said Ukraine struck the town 18 times with artillery fire and carried out more than 120 drone attacks on it and the main roads around it in the last week. Reuters reports.

SpaceX CEO Elon Musk on Saturday confirmed SpaceX restricted Ukraine’s Starlink internet service and accused Ukraine of violating SpaceX’s terms of service by using Starlink for military activity, but denied being responsible for a simultaneous issue with its U.S.-run counterpart Starshield. Ukrainian Presidential Advisor Serhii Beskrestnov said Musk’s post was “very strange.” Ukrainian Defense Minister Yevhenii Khmara earlier said that the military was now recording a “stable connection.” Aleks Phillips reports for BBC News.

SUDANESE CIVIL WAR

Sudan’s state electricity company today said Rapid Support Forces drone strikes hit power stations in four states in southeast Sudan. The RSF did not immediately respond to a request for comment. Khalid Abdelaziz reports for Reuters.

The United Kingdom today sanctioned 15 Sudanese military commanders, militia leaders, and security officials from both sides of Sudan’s conflict, accusing them of waging sexual violence, torture, and attacks on civilians linked to the fall of the North Darfur capital al-Fashir. The United Kingdom said the action followed the publication of U.K.-funded research by the Center for Information Resilience, which documented what it described as systematic starvation and conflict-related sexual violence against civilians by both sides in the war. Sam Tabahriti reports for Reuters.

OTHER GLOBAL DEVELOPMENTS

A joint statement by the United Kingdom and Canada yesterday confirmed the United Kingdom will not join the Canada-led Defense, Security, and Resilience Bank. The two countries instead said they would develop a framework for cooperation between their respective defense-funding initiatives, establishing two distinct but complementary institutions to support each other on investment opportunities and strategic challenges. Anusha Shah, Nivedita Balu, William Schomberg, and Elizabeth Piper report for Reuters.

Serbia’s long-standing leader Aleksandar Vučić yesterday admitted defeat to a student-led protest movement in snap parliamentary elections which were called after two years of protests, which were triggered by the collapse of a canopy in front of Novi Sad station that killed 16 people. Preliminary figures gave the students 2.2 million votes and Vučić’s Serbian Progressive Party 1.72 million votes. Vučić spent two terms as president and was leading his party with the aim of returning to the post of prime minister. André Rhoden-Paul and Jaroslav Lukiv report for BBC News.

Brazilian President Luiz Inácio Lula da Silva has been re-elected after yesterday’s second round vote-runoff against Sen. Flávio Bolsonaro, who has not conceded. Brazil’s Superior Electoral Court announced Lula had 51.79% of the votes. Bolsonaro refused to recognize the result, citing the extension of polling by two hours as “inexplicable,” despite both candidate’s campaign teams requesting the polling stations be kept open longer, and told supporters in the capital Brasilia that he would seek the annulment of the election. Lula and his supporters celebrated what they describe as a “historic victory” and a surprise turn-around. Marina Rossi, Leandro Prazeres, and Vanessa Buschschlüter report for BBC News.

Venezuelan opposition leader María Corina Machado yesterday returned to the country after nearly a year in exile. Two government officials told the New York Times Machado entered Venezuela from neighboring Colombia. The Trump administration previously made clear in private exchanges with foreign officials that it did not support Machado’s return, according to U.S. officials, but on this occasion U.S. officials decided not to intervene. Julie Turkewitz, Edward Wong, and Adam Entous report.

A criminal investigation into Argentinian President Javier Milei over links to a cryptocurrency scandal on Friday was dropped. Milei denied promoting the newly launched cryptocoin $Libra, which collapsed shortly after he mentioned it in a social media post in Feb. 2025. The case was dismissed by a federal judge at the request of prosecutors, who said that the “reported events do not fall under a legal category,” and concluded that no “criminally relevant deception” had been proven. Adam Goldsmith reports for BBC News.

North Korea last week sent a letter to the U.N. Security Council, repeating accusations that South Korea fabricated a September landmine incident. North Korean armoured vehicles were spotted today near the border by a Reuters witness at an observatory in South Korea overlooking the border. Hongji Kim, Kyu-Seok Shim, and Joyce Lee report.

CLIMATE DEVELOPMENTS

Director General of the U.N. International Atomic Energy Agency Rafael Grossi on Friday said the agency’s annual report to the U.N. General Assembly expects the use of nuclear power could more than triple by 2060. The growth will be driven by electrification, data centers powered by AI, and “technology companies’ appetite for reliable, carbon-free power.” Edith M. Lederer reports for AP News.

U.S. AI DEVELOPMENTS

The FBI is looking at potentially adding an economic espionage charge against Saudi national and former OpenAI employee Abdulrahman Alfozan, a source told the Wall Street Journal. The source said the investigation centers on the possibility that Alfozan stole information from OpenAI in order to offer it to the Saudi government, even if he wasn’t recruited to do so. A lawyer for Alfozan said he hadn’t received any information from the government on potential additional charges. Dave Lawler and Angel Au-Yeung report.

An Anthropic spokesperson yesterday said Claude accessed publicly available Georgia voter data without human authorization. Claude accessed aggregate vote counts while on an internal statistics project. The spokesperson said Anthropic has “since put in place mitigations to block these kinds of actions.” Maria Curi and Alex Isenstadt report for Axios.

Texas Attorney General Ken Paxton on Friday launched a lawsuit against OpenAI and the company’s CEO Sam Altman, alleging they misled consumers about the safety of the company’s AI chatbot and failed to protect users from harm. The suit alleges that OpenAI’s business practices violated the state’s consumer protection and children’s online safety laws. Paxton said that with the rapid advancement of AI, “OpenAI has the responsibility to protect users and warn them of the many dangers of its platform.” Casey He reports for POLITICO.

Political candidates have been engineering their campaign messages to appeal to the AI chatbot that more voters have been using to answer questions about issues, according to documents reviewed by Reuters. Mike Stone reports.

A U.S. Government Accountability Office report published on Oct. 14 observed that the Defense Department is building weapons and projects with AI capabilities that officials said don’t always closely adhere to the agency’s own guidelines for AI. When responding to the GAO, six of the ten officials surveyed said their teams had “identified risk(s)” with using the projects “in an operational environment.” Defense Secretary Pete Hegseth has repeatedly dismissed previous caution of AI tools and encouraged rapid autonomous innovation at the Pentagon. Chiara Eisner reports for NPR.

U.S. FOREIGN AFFAIRS

Sources told Reuters that U.S. administrations have been requesting SpaceX turn its Starlink service on or off in various countries for several years. During former President Biden’s term, the requests included switching off the service to prevent Iranian proxy militias in Yemen and Lebanon from using Starlink to communicate. Trump requested Musk make the service available in Iran following U.S.-Israeli strikes in 2025 and 2026, in hopes it would bolster internet access and communications for anti-government protestors. Erin Banco and Joey Roulette report.

Netanyahu in recent days pleaded with Trump to show more public support for his campaign, sources told Axios. The U.S. source said the White House sees Netanyahu’s aggressive lobbying as a “sign of desperation” and an indication that the Israeli leader thinks he’s trailing. Neta Molcho, a spokesperson for Netanyahu’s Likud party, denied that Netanyahu asked Trump to support his campaign “in recent days or before his endorsement statement.” Trump yesterday evening issued another statement endorsing Netanyahu, mainly because of his position on Iran. Barak Ravid reports.

The Wall Street Journal Editorial Board yesterday published an article where they detailed that “sources from three countries confirm that the Trump Administration has advocated Viktor Orbán for the role of United Nations Secretary-General.” The United States encouraged the former Hungarian prime minister to enter the race, and a State Department official said Russia and China are believed to be supportive, with the plan being that the three countries endorse Orban as their joint candidate.

Japanese prosecutors yesterday indicted U.S. Marine Lance Corporal Devin Jacob Ballard on charges of robbery and murder of Anna Yagi in a hotel in Okinawa prefecture’s capital Naha. A Japanese public broadcaster said Ballard allegedly strangled the victim with a telephone cord and stole a backpack containing a wallet. Satoshi Sugiyama reports for Reuters.

Trump yesterday at a political rally in Port Huron, Michigan predicted a trade dispute with Canada would be resolved “soon.” Trump said Canada is “willing to give us anything at this moment.” At the rally, Trump also called Norway a “shit house country” and said Norwegian immigrants were not welcome in the United States because the Nobel committee, based in Norway’s capital Oslo, had not awarded him a peace prize. There was no indication that the Trump administration was adopting a new official policy on Norwegian immigrants. Valerie Volcovici reports for Reuters; Bo Erickson and Jeffrey Dastin report for Reuters.

U.S. IMMIGRATION DEVELOPMENTS

Amnesty International on Friday urged Sierra Leone to halt the expected repatriation of three people deported from the United States, saying they could face serious human rights violations if they returned to their home countries. The rights group said they had each lived in the United States for more than 20 years when they were removed by ICE, and that they all have children under-18 still in the United States. Ayen Deng Bior reports for Reuters.

A federal appeals court on Friday ruled 10-7 that ICE detainees being held in Texas have no rights to seek release on bond, even if they’ve been held for years while their immigration proceedings play out. The court concluded that Congress had intended to restrict the availability of release for people facing potential deportation, even if they have lived in the United States for years without criminal records and had attended their ICE check-ins. Kyle Cheney reports for POLITICO.

A federal appeals court on Friday ruled that Trump violated federal law by seizing control of the California National Guard last year and deploying troops to the streets of Los Angeles during immigration protests. The decision applied the reasoning the Supreme Court used when it ruled that Trump did not have the power to send National Guard troops into Chicago last year. Laurel Rosenhall reports for the New York Times.

U.S. DOMESTIC DEVELOPMENTS

A federal Court of Appeals on Friday allowed portions of a Mississippi ban on DEI practices in public schools to go into effect. The parts of the law in question prevent public schools from requiring diversity training and maintaining DEI-related programs, courses, and offices. It also bans public schools from discussing a list of “divisive concepts” related to race, sex, gender identity, sexual orientation, religion, ethnicity, and national origin. Public schools that violate the law could lose state funding. Sophie Bates reports for AP News.

The Catholic Archdiocese of New York on Friday reached an agreement to pay more than $800 million to settle claims from 1,300 people who said they had been sexually abused by its clergy members and lay leaders. In an email to the priests of the archdiocese shared with the New York Times, Archbishop Ronald A. Hicks on Saturday confirmed the agreement. Maya King and Jeffery C. Mays report.

The Justice Department last week charged former FBI officials Nikki Floris and Bradley Benavides, alleging they lied to Congress in 2023 about the bureau’s probe into Hunter Biden, including its handling of Hunter’s laptop in the run-up to the 2020 election. The pair worked on the FBI’s now-dissolved Foreign Influence Task Force, and each face one count of making a false statement to Congress. Casey He reports for POLITICO.

The DOJ on Friday announced that FBI investigators arrested government contractor Ian Brinksman and charged him with leaking classified national defense information. Brinksman is accused of leaking the information to his girlfriend, who writes an online publication on Substack. Ford McCracken reports for The Hill.

TRUMP ADMINISTRATION ACTION

The DOJ on Friday sent letters to local election officials across California demanding they help identify potential noncitizens on their voter rolls and set aside their ballots. A spokesman for the U.S. attorney’s office in Los Angeles told the New York Times that the letter also said local officials would “risk significant federal scrutiny” if they failed to comply with DOJ demands. A panel of federal appellate judges yesterday questioned the timing, constitutionality, and authority of the DOJ’s demand. Laurel Rosenhall and Shawn Hubler report; Rebecca Boone reports for AP News.

Catholic Archbishop James R. Golka of Denver yesterday told a local Catholic publication that it was never his intention for a prayer meeting between Trump and several Roman Catholic bishops to be used as an endorsement. Bishop William Joensen of Des Moines wrote afterward that “I wish I had been more shrewd and less simple, like those who exploited our meeting for partisan political purposes.” Mitch Smith reports for the New York Times.

Andrew Gradison, a senior adviser at Trump’s Veterans Health Administration, this month told a conference for psychedelics enthusiasts that he is eager to ramp up the VA’s capacity to give psychedelic drugs to as many as 2.1 million veterans with mental illnesses like PTSD or depression. No government regulator has deemed psychedelics safe or effective for mental illness. Amanda Friedman reports for POLITICO.

Trump yesterday said he is considering moving the White House press corps from its historic location. Trump said he will use the current White House press corps location “for very badly needed space,” but will not build “a swimming pool in the White House Briefing Room.” Mira Bhakta reports for The Hill.

TRUMP ADMINISTRATION LITIGATION

The DOJ on Thursday sued Alaska to delay certification of election results if incumbent Sen. Dan Sullivan (R) loses. The Republicans accused Democrats of trying to confuse voters by planting a candidate with the same name, and the suit argued that an election result may be “impacted by voter confusion if: (1) Senator Sullivan does not win the election; and (2) had all votes cast for Daniel J. Sullivan been cast instead for Senator Sullivan and Alaska’s rank-choice process played out accordingly, Senator Sullivan would have won the election.” Trump on Friday said he personally directed the DOJ to file the lawsuit. Avery Lotz reports for Axios; Ben Johansen reports for POLITICO.

Attorney General Todd Blanche yesterday defended the DOJ’s lawsuit in Alaska. Earlier in the week he had said the DOJ could get involved in the vote certification process in a “doomsday” scenario. Blanche denied the idea that the suit was setting the stage for Trump to challenge the outcome of the midterm elections, and framed the case as a test of whether Alaska had violated federal law and the Constitution by permitting a potentially confusing ballot process to continue. Willem Marx reports for NPR.

Did you miss this? Stay up-to-date with our Litigation Tracker: Legal Challenges to Trump Administration Actions.

If you enjoy listening, Just Security’s analytic articles are also available in audio form on the justsecurity.org website.

ICYMI: Last week on Just Security

After a $25 Billion Deception Ruling, Why Trust Meta with an AI Agent?

By Mariana Olaizola Rosenblat

The post Early Edition: October 26, 2026 appeared first on Just Security.

Abd al-Rahim al-Nashiri was able to keep his focus and for the most part responded appropriately through an Arabic language translator during a nearly five-hour meeting inside his prison building, Army Maj. Anthony Ralston, the doctor, testified Oct. 22 and 23. Ralston evaluated al-Nashiri’s mental competence as part of a Sanity Board ordered by Army Col. Matthew Fitzgerald, the judge overseeing the case. 

Al-Nashiri, a Saudi citizen, is accused of conspiring in the suicide bombing of the U.S. Navy warship in Aden Harbor, Yemen on Oct. 12, 2000. Seventeen U.S. sailors were killed in the al-Qaeda attack, and dozens more were wounded. Jury selection is set to start Nov. 2 if Fitzgerald finds him competent for trial.

Ralston offered a starkly different description of al-Nashiri’s cognitive abilities from testimony earlier in the week by a psychiatrist and psychologist, both Pentagon-paid defense experts. They declared him incapable of understanding the court proceedings and assisting his lawyers in his defense, making him incompetent.

Ralston quoted al-Nashiri as telling him several times “I’m not crazy” during his examination. Asked about his mood, the prisoner replied “garbage.” But the doctor said the prisoner’s “outward appearance of emotion, attention was more jovial, more engaging than someone who felt like garbage.”

Prosecutors called no witnesses, leaving it to Allison Miller, al-Nashiri’s lead lawyer, to call Ralston and question him first about his report. He explained that there is a very high standard for finding someone incompetent, and al-Nashiri didn’t qualify. He said he had previously conducted 13 competency exams in his five years as a military psychiatrist and found two prisoners incompetent.

“Just having a diagnosis of PTSD doesn’t make you incompetent,” Ralston said. Someone who is schizophrenic can be competent, he said, but someone who is floridly psychotic is “probably going to be incompetent.”

Al-Nashiri was held by the CIA from 2002 to 2006. At an agency black site in Thailand, two contract psychologists waterboarded him and held him naked in a dark cramped wooden box among other “enhanced interrogation techniques” in an effort to get him to disclose future planned attacks and other al-Qaeda secrets.

Ralston said he had “no doubt” the prisoner had PTSD and depression and also an understandable distrust of his U.S. medical providers at Guantanamo. So, the psychiatrist said, he stopped shaving days before meeting al-Nashiri to conduct the mental assessment and wore a Hawaiian shirt and tan pants to look less like a military physician.

It will now be up to Judge Fitzgerald to rule on his competency after closing arguments Monday. 

For the defense, Dr. Terry Kupers, a psychiatrist since 1972, met the prisoner for roughly the same amount of time as Ralston but over three days at the courthouse complex. Al-Nashiri has vomited on himself during the 10- to 15-minute commute from the prison to court, his lawyers say, because he has flashbacks to his time in CIA custody, and so he is medicated with an anti-nausea drug that makes him drowsy on the ride – and then is given a stimulant for ADHD, which make him anxious and sometimes act out.

Ralston, a psychiatrist since 2021, met al-Nashiri just once, for four hours and 44 minutes, no break, at the Camp 5 prison where he could walk from his cell to a conference room, no need for sedatives or stimulants. 

Al-Nashiri’s lawyers had asked Ralston to also meet the defendant a separate time at the courthouse complex, to see how his behaviors changed. But the prisoner refused a second meeting.

Kupers said the prisoner’s attention was so scattered at the courthouse during his examinations that the doctor abandoned the idea of conducting a baseline 20-question test that evaluates a defendant’s understanding of the proceedings and roles of different participants in a criminal trial.

Ralston said al-Nashiri had taken a Tylenol and migraine medication during their interview and would digress at times. But Ralston said he was able to get the prisoner back on topic and conduct basic testing, including asking him to memorize five words and interpret proverbs. 

Al-Nashiri was able to recall three of the words five minutes later. Through an Arabic translator he interpreted the expression “People in glass houses should not throw stones” to mean “If you are weak then don’t fight with others.”

Allison Miller, the prisoner’s capital defense lawyer, cited the reply as an example of the prisoner getting it wrong. Ralston called his interpretation “abstract but idiosyncratic.” 

Over the objections of prosecutor Navy Cmdr. Clayton Trivett, the judge also heard testimony from Dr. Sondra Crosby, a recognized expert on diagnosing and treating torture victims. She has estimated that in her Boston-based practice she has evaluated about 1,000 torture survivors and has been meeting with al-Nashiri at Guantanamo for 13 years.

Crosby is an internist and until recently served as a commander in the Navy Reserve. She was the first doctor to diagnose al-Nashiri with PTSD, a finding that was upheld by a military medical board in 2013. But she is not a mental health practitioner, and so unqualified to offer an opinion on a person’s competence to stand trial. 

Instead, Crosby described for the judge a series of episodes of violent and humiliating abuse that al-Nashiri experienced in CIA custody, including some episodes that were described in CIA reports and court testimony that al-Nashiri does not remember.

“He is one of the most damaged people that I have seen,” Crosby said. “He survived a brutal period of torture during 2002 to 2006 that was designed to break him. And it did.”

Crosby said al-Nashiri’s flashbacks and nightmares have been better and worse across the years and that he “cannot regulate himself.” He has angry outbursts over minor things and is known to bang loudly on his cell door over his grievances.

Concentration is also a problem, she said, and he complains of memory loss, although he does not consider himself to be ill.

“Not once in 13 years has he completed a task” assigned by Crosby, such as keeping a sleep journal and, separately, a record of his nightmares. “He’s very apologetic,” she said, but has never done it.

To the suggestion that al-Nashiri was malingering, Crosby replied: “There is no way that someone can malinger for 13 years and be as consistent.”

The post Guantanamo Dispatches: Army Doctor Diagnosed USS Cole Bombing Defendant with PTSD and Depression, Found Him Fit for Trial appeared first on Just Security.

Meta’s announcement of Muse was accompanied by promises similar to what the jury in New Mexico just found deceptive. Its official launch page states that “Meta built Muse from the ground up to be a safe, secure, private, and widely available personal AI agent.” The company claims that conversations users have with Muse and the data contained in the “virtual machine” (VM), the isolated environment that Meta promises for each Muse user, are not shared with the company’s advertising systems. Security vulnerabilities have already been reported, however, and early adopters say that Muse overstepped its data access permissions (Meta disputes this account). 

The Stakes are Higher

The stakes with Muse are much higher than they were with Facebook. The data that got exposed during the Cambridge Analytica scandal underlying the New Mexico case was standard Facebook profile data. Muse can look more deeply into people’s personal lives, as it connects to their inboxes, calendars, payment accounts, health services, and smart home systems. Muse’s personalization features and voice (which Meta calls a “soul”) foster a sense of companionship that is likely to lead users to reveal even more about themselves through conversations. 

And Meta is also bringing Muse to its AI glasses starting next month, which could expose even more granular and intimate data. As I wrote previously, glasses equipped with sensors can capture what someone looks at, how long they linger, and how their body responds — body-based data that a person cannot review or withhold, often because they don’t know they are sharing it.

But The Protection of User Data is Narrow

Meta’s statement about keeping Muse data from its ad systems is narrower than it seems—but it takes reading the fine print to realize that. Without clarification, the statement could even be considered misleading. 

First, neither Muse’s launch page nor its privacy policy clearly defines what data is stored in the virtual machine. Muse’s launch page says the VM contains “the data and credentials for any service a person connects” to Muse. The privacy policy’s wording is even more ambiguous, referring to the VM as “the core environment where Muse keeps information you share and things you create together.” Yet Meta collects a range of data from Muse users—such as “information about how you use Muse, including features used, websites visited, actions and transactions, timing and frequency, and associated metadata”—that isn’t explicitly covered by the pledge not to share consumer data with its ad system. In other words, the data Meta is collecting is not necessarily stored in the VM, and that means Meta may be free to share it with advertisers. But potential users could read the promise on the launch page and incorrectly assume that all their Muse-related data will be withheld from advertisers. 

Second, Meta can still access the data in the VM, and the company may one day simply decide to change its privacy policy (as it has done many times in the past) to monetize some of that data. 

Given the New Mexico verdict and other precedents, one can easily imagine Meta using these ambiguities and loopholes to supplement its revenue. Advertising, after all, is the company’s business model. In last week’s earnings call, the company acknowledged Muse comes with “substantial compute costs” and gestured at the possibility of using Muse to improve personalization and advertising across Meta’s apps.

A company that has such a well-documented record of deceiving users—particularly where its use of customer data is concerned—should be as clear and transparent as possible about its intentions to monetize user data related to Muse, now and in the future. But, as we have argued repeatedly at the Stern Center for Business and Human Rights (where I work), protection of users’ rights cannot be left to companies alone, or to courts to remedy after the fact. The U.S. government must ensure a robust federal capacity to oversee products like Muse before more harm occurs, starting with additional resources for the Federal Trade Commission’s Office of Technology and comprehensive data privacy regulation. In the meantime, after the New Mexico verdict, consumers should be wary that using Muse means entrusting Meta with even more sensitive data.

The post After a $25 Billion Deception Ruling, Why Trust Meta with an AI Agent? appeared first on Just Security.

From its founding in 1866, the Klan and groups like it waged a campaign of terror across the South. In Louisiana alone, they killed roughly 1,000 people, most of them Black, in the months before the 1868 election. Targets were Black men who voted or held office, teachers in freedmen’s schools, families who acquired land or livestock, and anyone who refused the deference white Southerners demanded. The purpose was to restore by force the racial order that emancipation and the Reconstruction Amendments had formally abolished.

The Republicans who controlled the federal government responded with vigorous enforcement. Congress passed the Enforcement Acts of 1870 and 1871, making it a federal crime to conspire to deprive citizens of their constitutional rights and authorizing the president to use the military and suspend the writ of habeas corpus against organized terror. Congress created the Department of Justice in 1870, and President Grant appointed Amos Akerman as the Attorney General, who made subduing the Klan his central task. In October 1871, to prevent biased judges from releasing the perpetrators before they could be tried, Grant suspended habeas corpus in nine South Carolina counties, including York. Akerman traveled to oversee the crackdown himself. Federal marshals and soldiers arrested hundreds of Klansmen, and as many as 2,000 others fled the state.

What the historian Eric Foner calls “the legal offensive of 1871” reached across the South. Federal prosecutors indicted 220 Klan leaders in South Carolina. They indicted hundreds of men in North Carolina and nearly 700 in Mississippi. By the end of 1871, prosecutors had brought more than 3,000 indictments against Klan members and won more than 1,000 convictions.

By 1872, the federal government had broken the Klan. Racial terrorism dropped sharply. The 1872 election was the most peaceful of Reconstruction, with half a million Black voters casting ballots.

Within a few years, Northern political will ran out. Federal troops left the South, the Supreme Court gutted the Enforcement Acts, and the former Confederate states ended Black enfranchisement and reestablished white rule. But for a moment, the country saw what real enforcement could accomplish. 

Today, President Donald Trump understands the potential of federal enforcement. He has used it more aggressively than any modern president—but lawlessly. He has directed obliging DOJ attorneys to prosecute his political opponents, stripped universities and nonprofits of billions in funding based on unsupported legal claims, and punished law firms and media outlets for First Amendment activity. At the same time, his administration has dismantled enforcement that protects ordinary people from the powerful—gutting the Consumer Financial Protection Bureau (CFPB), contorting DOJ’s Civil Rights Division to target diversity programs and suppress the vote, and sidelining the Foreign Corrupt Practices Act.

When Trump leaves office, the next administration might approach enforcement warily, in an attempt to distance itself from this period of flagrant abuse. But overcorrection would be a mistake. The right response is to reject the current administration’s mob tactics and vendetta-driven prosecutions, while leaning into enforcement as a strategy to advance fairness, justice, and the rule of law. A range of federal statutes—across civil rights, consumer protection, labor rights, antitrust, banking, tax, and environmental law—exist for these purposes.

We need more of this kind of enforcement, not less, against those who use their power to cheat, exploit, dominate, and discriminate against others in violation of U.S. law.

The Problem: Decades of Neglect

Before Trump and outside the context of immigration, recent presidents of both parties have mostly treated enforcement as a technical function, something to leave to the agencies while the White House focuses on other policy strategies for advancing social and economic inclusion—legislation, rulemaking, grants, and benefits delivery. As a result, enforcement has lacked the political direction, interagency coordination, and public communications priority that other policy approaches receive. The costs have been significant.

During the financial crisis, the Obama administration bailed out the banks, let executives keep their bonuses, and sent only a single Wall Street banker to jail while millions of victims of predatory lending lost their homes in foreclosure. For four decades, the antitrust agencies stood aside as big corporations consolidated industry after industry. The National Labor Relations Board (NLRB) long underused its power to seek temporary injunctions during organizing drives, for example to reinstate a fired worker while the Board decides the case. Employers are charged with breaking federal law in more than 40% of union elections and with illegally firing workers in one of every five union election campaigns.

The causes of such neglect vary, from competing priorities and concerns about the appearance of political interference to industry capture and decisions to narrow and underfund enforcement. 

Presidents might see enforcement as small compared to, say, passing new legislation or strengthening the social safety net. Fair enough. But without strong enforcement, statutory protections are just words on a page. Moreover, enforcement has unique advantages, such as the potential to deliver tangible results quickly and to show the American people that their government is fighting for them.

Presidents have also kept their distance to avoid the suggestion of improper influence. Most modern administrations, including Trump’s first, have restricted White House contact with enforcement agencies to keep politics out of investigative and charging decisions. That policy makes sense when it comes to specific criminal and civil cases. But over time, the policy has served to box presidents out of enforcement strategy too, as if deciding which harms deserve the government’s attention were as dangerous as deciding whom to prosecute. That’s a policy matter and in fact a core presidential responsibility. 

Meanwhile, some agencies have absorbed a deregulatory culture that treats vigorous enforcement as hostile to growth. Recall the now-abolished Office of Thrift Supervision, whose director posed with a chainsaw beside a stack of bank regulations a few years before the financial crisis. Agencies have been weakened by revolving-door hiring practices or captured by the industries they oversee.

Presidents and Congress have also made specific choices that have constrained enforcement of civil rights and economic justice statutes. Beginning in the Reagan administration, the leaders of the Civil Rights Division and the Equal Employment Opportunity Commission (EEOC) moved away from strategic cases with wide impact and toward processing individual complaints. Over the years, Congress starved the agencies of resources. As the political scientist Robert Lieberman has documented in a forthcoming paper for the ANNALS, inflation-adjusted funding for the six major federal civil rights agencies was flat between 1981 and 2024—about $1 billion in 2024 dollars in both years—while the agencies were given more statutes to enforce and the number of people they serve (employees, students, tenants, patients, etc.) grew enormously. For example, the EEOC’s staff fell by a third over that period, even as the agency added the Americans with Disabilities Act and other laws to its mandate and as total employment grew by roughly 60%.

Yet history offers useful models of robust enforcement since Reconstruction. In 1938, President Franklin Delano Roosevelt (FDR) took up an aggressive antimonopoly agenda, arguing that excessive corporate concentration and private economic power threatened democracy. His appointee to lead DOJ’s Antitrust Division, Thurman Arnold, quintupled the number of antitrust lawyers on staff and brought more cases in three years than the government had in the previous half century (though World War II cut these efforts short). 

In the 1960s, the federal government intervened throughout the South to integrate public schools and protect Black voters through lawsuits, threats to cut off federal funds, and federal troops. The share of Black students attending majority-white schools in the South rose from 2% in 1964 to 33% in 1970. In Mississippi, Black voter registration rose from less than 7% to almost 60% between 1964 and 1967, as federal examiners bypassed local registrars. This period also saw landmark cases in employment discrimination. In 1973, the EEOC, the Labor Department, and the Justice Department forced AT&T, then the nation’s largest private employer, to pay $15 million to 13,000 women and 2,000 men of color and raise pay for 36,000 workers held back by discrimination. A year later, a consent decree with nine steel companies and the steelworkers union delivered $31 million in back pay and opened better jobs to women and workers of color across the industry. 

These efforts had presidential commitment, resources that matched the ambition, and strategies aimed at systemic change.

The next administration will have to rebuild under dire conditions. An estimated 70% of the Civil Rights Division’s lawyers have left, many replaced by people selected to execute the current leadership’s goals of undermining civil rights. The Education Department issued layoff notices to all but 60 of the 575 employees in its Office for Civil Rights and closed more than half its regional offices. The EEOC’s staff is at its lowest level since at least 1981. The current administration has tried repeatedly to fire everyone at the CFPB and shut the agency down. Many of the career staff who remain are demoralized. 

In addition, Americans who have watched Trump turn enforcement into a weapon could be skeptical of aggressive approaches from the next president. Finally, future enforcement targets, such as companies that cut pay-to-play deals with the current administration, might allege that they are the victims of political payback.

The Goals: Four Principles for Enforcement

A new approach to enforcement should serve four ends.

First, enforcement should shift power, not just remedy harm. After all, the laws at the center of this agenda were written to advance equality, dignity, and opportunity by correcting power imbalances that had enabled abuse, exploitation, and discrimination. For example, the National Labor Relations Act names the “inequality of bargaining power” between workers and employers as the problem it exists to fix. Antidiscrimination laws were passed to change oppressive political and economic dynamics, by arming people of color, women, people with disabilities, and others with new legal rights to assert. When violators can count on sporadic enforcement or absorb penalties as a cost of doing business, nothing really changes. The best test of a successful case is whether the party that abused its position can still do so afterward. Stopping one scheme is good. Removing the market dominance, undue positional leverage, litigation capacity advantage, and information control that made the abuse possible is better.

Second, enforcement should make rights real in ways people can see. When rights go unenforced, it tells people that the law is a game, or worse, that it belongs to someone else. Enforcement that puts money back in workers’ pockets, opens lending in redlined neighborhoods, and holds executives personally accountable demonstrates that the law has meaning.

Third, enforcement should be distributed and durable. Federal enforcement has proven too easy to switch off. More enforcement capacity should rest with state attorneys general, private litigants, and organized communities that have a role in implementing remedies.

Fourth, enforcement should be trustworthy. Presidents should set priorities—in public. But investigators and prosecutors should decide which charges or civil claims to pursue based on the facts and the law. And there should be clear protocols in place to ensure White House officials do not direct selective actions against the president’s perceived enemies or undue lenience for the president’s allies.

The Solution: Muscular and Fair Enforcement

The next administration will not have the luxury of rebuilding the old enforcement system before putting it to work. Agencies will have fewer real experts on staff, a good many people who never should have been hired, data and workflow systems that were sabotaged to enable politicized action, and a sullied reputation. But an administration that sets clear priorities in the public interest, coordinates across agencies, chooses cases for maximum impact, demands strong remedies, and explains its work in resonant plain language can accomplish a great deal even in these circumstances. 

Executive Action

Set an Enforcement Agenda. Presidents have coordinated rulemaking priorities from the White House for nearly half a century. Jimmy Carter first required agencies to publish notice and provide the White House analyses of their significant planned rules in 1978, a process that evolved into the modern Unified Regulatory Agenda. Ronald Reagan centralized review, effectively giving the Office of Information and Regulatory Affairs (OIRA) the power to stop agency proposals and using it to drive deregulation. Bill Clinton kept that machinery, used it to pursue his goals, and enhanced presidential influence by issuing directives to regulatory agencies and announcing proposed and final rules as his own. Presidential leadership over the regulatory agenda has only increased since then.

The next president should take a page out of the regulatory book and launch an Enforcement Agenda in the first days of the administration. That agenda should name the harms the government will go after, defined by conduct and sector: for example, price gouging, including based on AI-enhanced surveillance; wage theft and employee misclassification in high-violation industries; discrimination in housing and lending; monopolization in health care; and public corruption. Doing so would tell every agency that such enforcement is a presidential priority and set expectations for results. Agencies would make responsive department-level commitments immediately. They would be expected to map their authorities to the president’s enforcement priorities, design practical implementation strategies (e.g., intergovernmental strike forces, portals to gather evidence from the public, whistleblower protections), and allocate resources accordingly. 

This is not a call to replicate the cumbersome bureaucratic processes of regulatory review or to expand the mandate of OIRA, which has no enforcement expertise. Nor should enforcement be directly “run out of the White House,” as Vice President JD Vance said of the work of the new Assistant Attorney General for nationwide fraud enforcement, operating “under the supervision of me and the President of the United States.” Notwithstanding the metastasis of the unitary executive theory, the president should not order specific prosecutions, whether face-to-face or by social media blast. Decisions about whom to investigate and charge should stay with agency personnel, rooted in apolitical application of the law to the facts.

The point is to drive more strategic and robust use of enforcement throughout the government, in service of struggling Americans. As legal scholar Kate Andrias has written, “presidential oversight of administrative enforcement is a core constitutional responsibility and has roots in early historical practice.” And there are several ways to catalyze agency commitments without politicizing enforcement. The president could issue an executive order laying out the Enforcement Agenda and requiring agencies to assemble implementation strategies by a given deadline. The president could issue specific directives to each agency, as President Biden did with his 2021 competition order. That order predated the Supreme Court’s ruling in Trump v. United States that the president has “exclusive authority over the investigative and prosecutorial functions of the Justice Department” and its ruling in Trump v. Slaughter bringing independent agencies under the president’s control, so Biden used circumspect language (e.g., agencies “are encouraged to enforce the antitrust laws”). But his order set out his priorities in detail, naming industries with a record of anti-competitive conduct and charging agencies to collect evidence, document legal violations, and pursue enforcement. Another approach is to create non-White House task forces but with more robust coordination, meeting, and reporting obligations. The Attorney General could lead, as in President Obama’s 2009 financial fraud task force. Agencies with a track record of aggressive enforcement, like recent incarnations of the CFPB and Federal Trade Commission (FTC), could also be good choices. 

Whatever the model, presidential prioritization would put a premium on speed, ambition, impact, and interagency cooperation. Agencies should be directed to collaborate to leverage the full authority of the government against corrosive practices. The same industry might be rampant with wage theft, price fixing, consumer scams, discriminatory hiring, and union busting. Traditionally, those practices likely would be investigated by four or five different agencies, or not at all. Under an Enforcement Agenda, they would plan together, share evidence as permitted under law, and rely on each other’s authorities to collectively advance the public interest. For agencies short on staff, or whose statutes have been eviscerated by the Supreme Court, pooling efforts this way will be a necessity.

Agencies would report consistently to the president, and the public, on their progress. Meanwhile, White House staff would work to empower, not restrain, enforcers. The policy councils and OMB would need to prioritize enforcement offices when developing the president’s budget and negotiating with appropriators. They would champion legislation expanding those offices’ jurisdiction and giving them stronger investigative and remedial tools. And the communications team would elevate enforcement wins for workers, consumers, and families from the White House, rather than leaving them to the smaller platform of agency public affairs offices.

Pursue cases that can provoke systemic change. With fewer investigators, agencies cannot afford to spend their time processing complaints one at a time. They should choose cases for their deterrent reach. During the Obama administration, the Labor Department’s Wage and Hour Division (WHD) went after the fast-food chains, hotel companies, and general contractors whose business models pushed cost pressure onto franchisees and subcontractors and bred wage theft at the bottom. WHD sought to hold those companies responsible for the violations they caused throughout their networks under joint employment liability. The agency also shifted toward proactive investigations in the industries with the worst violation rates. The idea was that one case against a large and egregious law-breaking firm could change practices at thousands of workplaces. Proactive investigations grew from 24% of WHD’s docket in 2008 to over 50% by 2017.

Agencies can also use data to surface violations that victims may not even be aware of. For example, the Biden administration’s Combating Redlining Initiative ran analytics on mortgage lending data to identify lenders that were avoiding Black and Latino neighborhoods, in violation of fair housing and fair lending laws. The Justice Department worked with the CFPB and other banking regulators to pursue a slate of cases, achieving 16 resolutions in less than four years—more than in the previous quarter century—and securing $153 million in relief for communities of color that had suffered discrimination. 

Use every tool already on the books. Agencies hold authorities they have allowed to lie dormant. During the Biden administration, the NLRB revived a doctrine from 1949 to make bargaining orders the default remedy when employers break the law during union campaigns. The Department of Labor leveraged the Fair Labor Standards Act’s “hot goods” provision, which lets the government stop the shipment of goods made in violation of labor rules, to crack down on child labor. Biden’s appointees to the FTC revived the Robinson-Patman Act’s ban on suppliers charging dominant retailers less than their smaller competitors. 

To make the most of limited time and resources, an incoming administration should have an inventory of every enforcement agency’s underused authorities before it takes office. Congressional committees of jurisdiction, former agency officials, and other legal experts can help catalogue these authorities over the next two years. 

Demand structural remedies. Too often, enforcement ends with a company promising to behave better. When the Justice Department approved Live Nation’s merger with Ticketmaster in 2010, it relied on the company’s commitment not to retaliate against venues that used rival ticketing services. Nine years later, the Department found Live Nation had repeatedly broken that commitment, and in 2024 DOJ and state attorneys general sued the company for antitrust violations. Behavioral remedies leave the underlying power in place, require years of monitoring that agencies cannot sustain, and eventually expire. In September 2026, twelve states settled their challenge to Paramount’s takeover of Warner Bros. Discovery, accepting only commitments of this kind, most lasting five years. The combined company will keep its size and reach long after those commitments lapse. 

When a company repeatedly or egregiously abuses its position, the government should seek structural remedies that curtail or eliminate its ability to continue violating the law. In severe cases, this should include breaking up monopolies through divestiture, revoking a bank charter or other business license, and debarment from federal contracting. Remedies should also empower those outside government to prevent harm from recurring in the future. Consent decrees should require public reporting of data on hiring, lending, pricing, and so on, so that workers, advocates, and journalists can hold companies to their obligations after the government moves on.

Name bad actors and show the public their government is fighting for them. An Enforcement Agenda should not direct agencies to go after specific targets in advance of fact gathering. That’s vital to the rule of law and to public confidence that enforcement is not payback. But once the evidence is in, the government should name names, loudly. The president and agency heads should explain in plain language which companies stole from which workers, which lenders shut which neighborhoods out of credit, and which monopolies squeezed which families. 

Last month, New York City Mayor Zohran Mamdani announced a record $131.5 million settlement with DoorDash for more than 260,000 delivery workers. He cited the DoorDash CEO’s own words that companies should “take the greedy algorithm and keep going,” and declared, “City Hall will not sit idly by while a megacorporation that made nearly a billion dollars in profits last year rips off the working people who keep our city moving.” The city’s consumer and worker protection agency took the enforcement action, and the mayor, appropriately, told New Yorkers what it meant for them. 

In fairness, federal enforcers should exercise caution when a case is still pending, to avoid prejudicing a jury pool or disclosing information that could compromise the integrity of the case. But the idea that federal enforcement agencies should speak only through their court filings, or that technical press releases with bloodless allusions to the rule of law will do the job, ignores how the modern media environment works. Effective enforcement has to break through. Moreover, people who have felt the outrage of being ripped off or the sting of discrimination deserve to experience vindication just as deeply.

Communication of this kind can deter others in the industry and rebuild public faith that the law protects ordinary people. It is also the best answer to the payback charge. Retaliation begins with an enemy and goes looking for a violation. Legitimate enforcement begins with the violation and shows the public who was harmed: the dishwasher shorted her pay, the family denied a mortgage, the patient overcharged for insulin. An administration that announces its priorities in advance, applies them across an industry, calls out the perpetrators, and puts the victims at the center of its story will make that difference plain. 

Rebuild the federal enforcement workforce, and collaborate with state and private enforcers. The administration should use direct hire, reinstatement, excepted service, and other rapid hiring mechanisms to bring back experienced career staff and recruit new enforcement personnel quickly. The legal nonprofit Governing for Impact recently inventoried these authorities, many of which are well tailored for surge enforcement hiring. For example, federal law allows agencies to expedite hiring into the competitive service when “there is a critical hiring need” occasioned by an “unanticipated or unusual event or mission requirement” or “to conform to the requirements of law, a Presidential directive or Administration initiative.” Restaffing gutted civil rights and consumer protection offices to enforce neglected statutes and implement a new president’s Enforcement Agenda would seem to qualify. The authority to reinstate prior employees would permit rapid onboarding of enforcers who felt forced to leave. The Intergovernmental Personnel Act allows agencies to bring on experts from state and local government as well as nonprofits for up to four years. Attorneys, meanwhile, are exempt from the standard competitive examination process and can be hired quickly.

The severe capacity deficit facing a new administration should also awaken it to the need to partner more fully with state attorneys general, city attorneys, district attorneys, and the private bar. These actors possess overlapping authority with federal agencies, and several states have increased their resources for enforcement in arenas abandoned by the current administration. The federal government should collaborate with them through joint investigations and lawsuits, evidence sharing, and amicus support for meritorious private cases. Indeed, such collaboration should become a standard federal strategy to magnify impact, not just a stopgap. Cases brought by states, local governments, and private plaintiffs also have a virtue federal cases lack: a future administration cannot simply drop them.

Congressional Action

Provide missing tools. An agency can enforce only the laws Congress has passed and conferred to its jurisdiction. Many of those authorities are missing, incomplete, or toothless, and many have been hollowed out by the courts or Congress itself. The Supreme Court has narrowed federal bribery law so much that officials can take payments for official acts with little fear of prosecution. There is no federal equivalent of Section 1983 for constitutional violations by federal officers. The NLRB can’t impose civil penalties, Title VII’s damages caps have not moved since 1991, and the FTC lost its power to recover money for defrauded consumers in 2021. Congress should close each of these gaps—potentially in one omnibus enforcement statute, given that the window for new legislation will likely be narrow—and expand authority for structural remedies.

Make enforcement harder to switch off. Congress has already authorized states to enforce over twenty federal statutes, such as Dodd-Frank’s consumer finance protections, the Clayton Act, the Clean Air Act, and the Clean Water Act. It should extend that model to protect workers, defend civil rights, and stop corruption. As an example, the recently introduced Anti-Corruption Bureau Creation Act would empower state attorneys general to sue the president, senior government officials, and federal contractors when officials use their positions to enrich themselves. Congress should also stimulate private enforcement of a range of federal statutes by reversing court-imposed limits on private rights of action and ending forced arbitration of civil rights and employment claims.

Fund enforcement at the scale of the problem. Immigration and Customs Enforcement (ICE) had an annual budget of $10 billion in fiscal year 2025. Since Trump took office, Congress has given ICE $113 billion more. Federal civil rights, labor rights, and consumer protection agencies combined are a fraction of ICE’s pre-Trump budget. A Congress that lavishes funds on an agency that has deported and killed American citizens can find the money to ensure workers are paid what they’re owed and families aren’t cheated out of their homes. Congress should multiply these budgets several times over and tie the new money to the kind of enforcement described here: strategic, coordinated, and aimed at the abuses that hurt Americans the most.

Create stronger guardrails against abuse. The headwinds facing an aggressive enforcement agenda will be fierce. In particular, cases involving the corruption being perpetrated now or involving companies that have allied themselves politically with the current administration are likely to be characterized in the distorted terms of payback and retaliation. To promote public faith in legitimate enforcement, and to prevent abuse, Congress should legislate new guardrails against politicized enforcement. Potential reforms include easing the legal burden to establish malicious and selective prosecution claims; allowing expedited discovery into the government’s motivations upon a threshold showing that a defendant was targeted for protected speech; and reforming grand jury rules, such as by requiring the presentation of exculpatory evidence and routinely providing transcripts in full to a judge and in part to the defendant. 

Conclusion

Of the period immediately after emancipation, Frederick Douglass observed: “The law on the side of freedom is of great advantage only where there is power to make that law respected.” His words apply with continuing force today. A robust enforcement agenda, communicated publicly and wielded responsibly but aggressively, should be understood as vital to the rule of law and an America that works for everyone. This Blueprint illustrates how a new administration can leverage the force of its broad enforcement authority to make federal rights real, hold those who violate them accountable, and give the American people a reason to trust that their government’s institutions are working for them. The staff and funding shortages will take time to fix. But if the next president makes enforcement a priority, the reinvention can begin on day one.

The post Reinvention Blueprint No. 5: Reinvigorating Federal Enforcement to Fight for the Public appeared first on Just Security.

Now, the report says, the question is how to get onto a pathway with the “lowest possible” remaining overshoot. This route aims to limit peak warming to 1.7-1.8°C by mid-century, preserving a 50 percent possibility of reduction to 1.5°C by 2100. This would require cutting emissions by 18 percent by 2030, 45 percent by 2035, and 69 percent by 2040 (all against 2025 levels). Countries’ current pledges, by comparison, would cut emissions by just 5-8 percent by 2030 and 11-15 percent by 2035; these are predicted to lead to about 2.6°C of warming – and that assumes they fully deliver on those reductions. Getting onto the “lowest possible” overshoot pathway, therefore, requires significant further interventions.

The externally reviewed report was produced by a team of 40 international authors, led by UNEP-Copenhagen Climate Centre Chief Scientific Editor Anne Olhoff, and including leading scientists from major climate-modeling and policy research centers.

Where might policymakers, academics, journalists, and others go from here? Below is a non-exhaustive list of questions to consider in light of this week’s findings.

1. What are the legal implications of overshooting 1.5°C?

In the International Court of Justice (ICJ)’s 2025 Advisory Opinion, the Court held, among other things, that the 1.5°C benchmark had become the States’ primary goal under the Paris Agreement, which calls for aiming for “well below 2°C” while pursuing efforts toward 1.5°C. Regional and domestic courts have also been referencing this benchmark in legal frameworks for State obligations with respect to climate change.

Now, UNEP concludes that the 1.5°C target cannot be met without overshoot. What, then, is the target’s legal effect? Does the obligation to pursue 1.5°C convert to an obligation to minimize the overshoot, or to something else? Does each additional year or one-tenth of a degree (the meaningful unit of measurement identified in the report) give rise to an additional claim of non-compliance? These and other questions related to overshoot are already being interrogated by the international legal community (including by our friends at the American Branch of the International Law Association’s International Law Weekend today), and the latest UNEP report only makes the questions more urgent.

2. How will courts use the report’s new emissions benchmarks?

A number of “fair-share” or “rights-based” climate change cases, largely in European national or regional courts, have measured State and corporate obligations against what would be necessary to achieve 1.5°C. Two Dutch cases, for example, benchmarked to 43-45 percent cuts by 2030. UNEP now calls for steeper cuts tied to a later baseline, but also calls part of the gap “unbridgeable” – so depending on how this is read, it could help or hurt plaintiffs in such cases. It also raises the question of how courts will treat judgments built on old figures, e.g. if appealed or if plaintiffs seek enforcement. And how might this intersection of litigation and shifting science impact domestic political will to entertain climate cases at all – particularly amid some signs of a backlash against climate change litigation, like New Zealand’s August 2026 law barring tort claims for climate change damages?

3. Who pays (or should pay) for the overshoot years?

UNEP’s report describes a process for bringing temperatures back down by 2100, implying a cost that will largely fall on future generations. This adds an additional dimension to long-running debates about obligations and remedies for climate change, which already involve stakeholders including major historical emitters, the least-developed countries that bear the worst impacts of climate change despite being only incidental contributors, and private companies.

4. How could 320 gigatons of carbon removal be governed?

The “lowest possible” overshoot pathway depends on 320 gigatons of carbon removal this century; the report notes that the global ability to do so “is uncertain, because of socioeconomic and technological challenges to the scale-up of CO2 removal measures and technologies.” Even assuming the scientific dimension is solved, this also raises a host of complex governance questions, including:

• Who is responsible for verifying removal and ensuring that what’s removed stays put?
• How would removal regimes compete with existing land use rights? What happens when such projects compete with food production, housing, or Indigenous rights? How do ocean-based removal projects intersect with international governance there?
• How do steps toward the 320 gigatons of removal intersect with existing obligations, e.g. could such removal be counted toward a State’s national contribution, and how would they be prevented from replacing emissions cuts?

5. What are the consequences of U.S. exit from the UNFCCC?

The U.S. withdrawal from the U.N. Framework Convention on Climate Change (UNFCCC) formally takes effect on Feb. 27, 2027, and the report notes that the withdrawal (along with revised estimates of historical emissions) has already worsened projected warming. It also warns that an apparent narrowing of the distance between State pledges and policies actually reflects removal of the U.S. pledge rather than real progress. 

The United States stands out in the report, including in a subsection entitled “Emissions growth slowed down in the G20, with the United States of America being the major exception” and in a footnote noting the U.S. disagreement with the report, included at the request of the U.S. State Department. The choices of the United States – the world’s second-largest current emitter and largest historical emitter – will necessarily shape the global path. A related question, then, is what other paths might be available to reduce U.S. emissions, such as state and local activity or civil-society or private-sector leadership.

6. How are the Strait of Hormuz crisis and broader geopolitical instability shaping the energy transition? 

The report notes that the Strait of Hormuz closure materially affected energy trends, driving up the price of oil and causing States to seek alternative energy sources. The net impact on emissions is unclear, as some States turned to coal and fuel subsidies while others have cut demand or are accelerating development of renewables. More broadly, conflicts emit directly and drive trends that, as the Strait of Hormuz shows, can both increase and decrease emissions depending on the second-order behavior that follows. The report also notes that energy security is increasingly a driver of clean-energy development in its own right.

7. How does each tenth of a degree of warming translate to our lives?

“In August 2026,” the report says, “global temperatures tied the record for the hottest month ever observed.” Extreme weather shapes health, labor protections, housing and insurance, migration, and nearly every dimension of public life.

The report takes each one-tenth of a degree, and each year above 1.5°C, as meaningful units of change. And with good reason: each also means real impacts on humans living in a climate-inflected world.

The post U.N. Emissions Gap Report 2026: “The Point of Return” — Seven Questions to Consider Next appeared first on Just Security.

A curated weekday guide to major news and developments over the last 24 hours. Here’s today’s news:

RUSSIA-UKRAINE WAR

Russian President Vladimir Putin yesterday said that future talks on a settlement of the war in Ukraine should focus on security guarantees for both countries and an overall framework for security throughout Europe. Putin said the peace should be in its “final form, not as a result of a temporary ceasefire for some short period of time.” Vladimir Soldatkin reports for Reuters.

Trump on Wednesday called SpaceX CEO Elon Musk and asked him to lift restrictions on Starlink in Ukraine, U.S. officials told the Wall Street Journal. The sources said the system was working more normally yesterday shortly after their conversation. Trump told Musk he was frustrated the system’s speeds remained capped. Trump yesterday said Musk had restored Ukraine’s Starlink access but declined to comment when asked about if he had spoken to Musk about the matter. Alexander Ward and Lara Seligman report; Jacob Bogage and Kanishka Singh report for Reuters.

Kremlin spokesman Dmitry Peskov today told reporters that reports that U.S. citizen Stephen James Hubbard was released in exchange for the recent release of Russian war blogger Mikhail Zvinchuk from custody in Montenegro were false. Dmitry Antonov reports for Reuters.

A tanker carrying 40,000 metric tons of Russian diesel is heading towards New York, ship-tracking data viewed by Reuters showed today. The tanker is the first to be transferred to the United States after Trump announced a deal to import Russian diesel.

ISRAEL-HAMAS WAR

Israeli strikes and gunfire in Gaza yesterday killed at least four people, according to local health officials. In Gaza City, a strike on a car killed at least one person, another strike killed a Palestinian man on a bike in the southern city of Khan Younis, and a strike killed a man near Nuseirat refugee camp in central Gaza. Gunfire killed a woman near the northern Jabalia refugee camp. The Israeli military said it had struck Hamas militants in Gaza City, Khan Younis, and Nuseirat, but that it was not aware of the reported incident in Jabalia. Nidal al-Mughrabi reports for Reuters.

WEST BANK DEVELOPMENTS

The Israeli military today said they prevented a “terror attack against Israeli civilians” in the occupied West Bank city of Hebron when they shot and killed two men. The IDF said the two men were carrying guns. The office of Israeli Prime Minister Benjamin Netanyahu later instructed security forces to cordon off Hebron and prepare to fully close all of the West Bank. Shlomo Mor and Sam Mednick report for AP News.

U.S. Ambassador to Israel Mike Huckabee is questioning whether President Trump’s Board of Peace can deliver on the peace plan for Gaza without more planning and resources from the U.S. government, according to a cable the embassy yesterday sent to U.S. State Department leadership obtained by POLITICO. The cable was signed by Huckabee and expressed concern that neither the Board of Peace nor the U.S. embassy in Jerusalem have the personnel to move forward with plans to disarm and rebuild Gaza. Dianella Cheslow reports.

SUDANESE CIVIL WAR

The Sudanese Rapid Support Forces last year systematically destroyed el-Fasher’s healthcare system during their siege and subsequent takeover of the city, according to an investigation by BBC News. U.N. High Commissioner for Human Rights Volker Türk called the findings “deeply disturbing,” and said attacks by the RSF on civilian infrastructure were serious violations of international humanitarian law and may amount to war crimes. The RSF has rejected BBC News’ findings. Barbara Plett Usher and Mohamed Madi report.

CLIMATE DEVELOPMENTS

A pair of studies published today in the Nature Health journal say severe heat killed more than 50,000 people in European cities and towns this summer. Researchers said that the figure is three times as many people as would have died had manmade global warming not occurred. The studies are based on a combination of factors that influence heat-related health risks in particular locations, and the estimates are not based on official death reports. Raymond Zhang reports for the New York Times.

The International Union for Conservation of Nature’s Red List of Ecosystems yesterday announced that almost all of the world’s shallow-water coral reefs have now been classified as critically endangered. It was the group’s first global assessment of shallow-water reef ecosystems, and it said heat stress because of climate change is the overwhelming reason for the finding. Catrin Einhorn reports for the New York Times.

OTHER GLOBAL DEVELOPMENTS

Advocacy group the Movement for an Open Web yesterday lobbied the EU and United Kingdom to suspend Meta Platforms’ new commercial terms due to come into effect on Oct. 30, alleging it would give the company unlimited right to use their content for its AI services. The group said it complained to the European Commission and the U.K. Competition and Markets Authority, and requested interim measures. Foo Yun Chee reports for Reuters.

Foreign diplomats in Brazil are discussing a rapid recognition of whoever wins Sunday’s presidential runoff, four ambassadors told Reuters. The ambassadors seek to ward off any attempts from outside the country to call the election prematurely for either candidate, but they have yet to formalize a plan. Brendan O’Boyle reports.

South Korea today proposed military talks with North Korea to resolve disagreements about where their border lies and prevent accidental clashes in frontline areas. It is unclear whether North Korea will agree to the talks. Kim Tong-Hyung reports for AP News.

U.S. FOREIGN AFFAIRS

South Korean and U.S. officials see slim chances of a meeting between Trump and North Korean leader Kim Jong Un this year, sources told Reuters. The sources said the United States has received no response from North Korea about the potential meeting, and there is little time left before a regional summit in China next month where the meeting might occur. Trevor Hunnicutt, David Brunnstrom, Antoni Slodkowski, and Ju-min Park report.

Global Affairs Canada, which manages the government’s diplomatic relations, on Tuesday issued guidance to Canadian diplomats urging them to avoid engaging with midterm election candidates, a day before the U.S. Embassy in Ottawa called out several Canadian leaders for political interference. The directive instructs officials to avoid being photographed with U.S. lawmakers and is in effect until Nov. 3. A Global Affairs Canada spokesperson said “government officials are always instructed to limit public events and stakeholder engagement during election periods in their host country.” Mickey Djuric, Nick Taylor-Vaisey, and Megan Messerly report for POLITICO.

The Treasury Department yesterday proposed a rule to crack down on U.S. high-tech investments such as supercomputing and AI in China, Russia, North Korea, and other “countries of concern.” The rule would expand current restrictions on investments that could help boost adversary military power to include investments in hypersonic systems capable of traveling at five times the speed of sound, supercomputing, biotechnology, and AI. Paul Wiseman and Didi Tang report for AP News.

U.S. IMMIGRATION DEVELOPMENTS

Minnesota state officials in October submitted court filings that said the federal government has stopped sharing evidence for investigations into the killings of Renee Good and Alex Pretti by ICE agents. State prosecutors announced in July that the federal government had begun turning over evidence in the cases, and Hennepin County Attorney Mary Moriarty said at the time the state expected to receive “all the evidence.” However the U.S. Attorney’s Office for Minnesota on Sep. 10 told Moriarty that evidence sharing would be on pause indefinitely “in light of challenges with D.C.,” according to the documents. Hannah Schoenbaum reports for AP News.

U.S. DOMESTIC DEVELOPMENTS

A report published on Monday by a cybersecurity forensic firm hired by Bard College to investigate whether their former president Dr. Leon Botstein sent an email to convicted sex offender Jeffrey Epstein suggested the alleged email could be fake. Dr. Botstein retired from his presidency over the summer and stepped away from all Bard positions this month amid allegations that he sent an email to Epstein in 2014 asking if a young foreign model was “for me, or are you keeping her for yourself.” Dr. Botstein welcomed the report and said “as I previously said, I did not author or send that email,” and that he appreciates “the effort to find the truth.” Vimal Patel reports for the New York Times.

Calley Means yesterday resigned from his role as a senior adviser for MAHA. Means expressed “tremendous respect” for Trump and Health Secretary Robert F. Kennedy Jr. and praised their efforts to take on “entrenched, dark forces in DC.” Adriel Bettelheim reports for Axios.

Defense Secretary Pete Hegseth’s team dropped and damaged a historic World War II-era globe, a source told The Hill yesterday. The globe, previously used by Gen. Dwight D. Eisenhower, was damaged while it was being moved to the secretary’s office. Filip Timotjia and Suryatapa Chakraborty report.

TRUMP ADMINISTRATION ACTION

Hegseth on Tuesday issued a new directive telling troops they cannot use “nonpublic information” to make bets on prediction markets, in a memo reviewed by AP News. Hegseth wrote that our “enemies watch movements in these markets,” and added that when a U.S. official “engages in a financial transaction on a military outcome, they create a signal.” Konstantin Toropin reports.

Trump yesterday encouraged Republican nominee for Ohio governor, Vivek Ramaswamy to continue running a controversial advertisement that includes the estranged mother of his opponent Dr. Amy Acton calling the Democrat candidate a liar. Trump said “Vivek should keep it up, and let everyone listen to the mother of a defective, dishonest, and incapable Candidate, explain how the daughter is a DISASTER who, uncontrollably, lies, cheats, and steals!” Trump criticized Ohio Gov. Mike DeWine (R) for urging Ramawamy to take down the ad and called him “disloyal to the Republican Party.” The ad involved Dr. Acton’s mother claiming Dr. Acton lied about her childhood, and while it does not explicitly reference Dr. Acton’s sexual assault, Dr. Acton’s stepfather was charged with raping the candidate when she was 12, and her mother fled from the state alongside him. Ashley Ahn reports for POLITICO.

Trump yesterday invoked the Defense Production Act and announced that the government would be spending $150 million toward boosting domestic gas and diesel production. The Cold War-era law gives the president broad powers to increase production of resources within the country for national defense purposes. The Energy Department in a press release noted that the incoming cash was “expected to increase U.S. refining output by millions of gallons per day.” Caroline Vakil reports for The Hill.

The Justice Department opened a criminal investigation into Corey Lewandowski over accusations that he profited from his role as a top adviser to the Department of Homeland Security, officials told the New York Times. The sources said the inquiry was initiated in response to a referral from the DHS’s inspector general. Trump yesterday said he had heard about the investigation but did not know the details. Glenn Thrush, Tyler Pager, and Nicholas Nehamas report.

The DOJ yesterday announced charges against Virginia state Sen. L Louise Lucas (D) for allegedly running a criminal enterprise that illegally bought and sold large amounts of marijuana and laundered the proceeds. Lucas was described as the ringleader of the group, which included her two daughters and grandson. Marijuana isn’t yet legal in Virginia. Sadie Gurman reports for the Wall Street Journal.

TRUMP ADMINISTRATION LITIGATION

The DOJ yesterday asked a federal appeals court to step in and accelerate a lawsuit against California which attempts to compel the state to not count as many as 20,000 mail-in ballots sent to voters the Trump administration claims are “likely noncitizens.” The request jumped ahead of a lower court that was still considering last week’s emergency request. Zach Montague reports for the New York Times.

The Supreme Court yesterday agreed that New Hampshire can direct election officials not to consider student IDs when verifying voters’ identities. The ruling did not include a vote count and explained that “lower federal courts should ordinarily not alter the election rules on the eve of an election.” Justice Ketanji Brown Jackson dissented, and wrote that she viewed the federal judge’s action as merely a “targeted order” to correct public statements and guidance by the states’ secretary of state that the lower court had concluded were “at best, misleading.” Abbie VanSickle reports for the New York Times.

A federal judge has ordered Trump to answer questions under oath in his defamation lawsuit against the BBC by Dec. 11. Trump is suing the BBC over an episode of Panorama that edited together two different sections of a speech he gave before the Jan. 6, 2021 riots. Trump previously canceled a deposition on Oct. 16 citing “matters of national importance,” but a lawyer representing the BBC said in court that Trump was campaigning in Georgia that day and had “prioritized that” over answering their questions. Bernd Debusmann Jr. reports.

Did you miss this? Stay up-to-date with our Litigation Tracker: Legal Challenges to Trump Administration Actions.

If you enjoy listening, Just Security’s analytic articles are also available in audio form on the justsecurity.org website.

ICYMI: Yesterday on Just Security

How Civil Society Can Help Unlock More Universal and Extraterritorial Jurisdiction Cases

By Courtney Spivey Urschel

The post Early Edition: October 23, 2026 appeared first on Just Security.

Par sa décision n° 2026-341 L du 15 octobre 2026, le Conseil constitutionnel redessine la part de la loi et celle du règlement dans la gouvernance des universités. Reprenant la définition des règles constitutives forgée en 1993, il n’y range plus que la détermination et le rôle des organes de direction et d’administration, le principe de leur élection ou de leur désignation et les grandes catégories de personnes qui y siègent. Le mode de scrutin, la durée des mandats, la voix prépondérante du président ou la part des professeurs au sein de la commission de la recherche deviennent ainsi des modalités d’organisation interne, qu’un décret en Conseil d’État pourra modifier. Parmi les dispositions dont le Premier ministre demandait le déclassement, seule la récusation devant la section disciplinaire des enseignants-chercheurs demeure dans la loi, parce qu’elle touche à l’impartialité d’une juridiction.

En 1982, le Conseil constitutionnel jugeait que la procédure de l’article 37, alinéa 2, de la Constitution a pour effet, « après la promulgation de la loi et par la voie d’un déclassement, de restituer l’exercice de son pouvoir réglementaire au Gouvernement[2] ». Son usage est longtemps resté mesuré. De 1959 à 2025, le Conseil a rendu en moyenne moins de cinq décisions de déclassement par an, sans qu’aucune année eût excédé dix décisions, et il a reconnu le caractère réglementaire des dispositions en cause dans plus de 70 % des cas[3]. Cet instrument, dont on a relevé qu’il a « vocation à être utilisé de manière ponctuelle[4] », a pourtant donné lieu depuis le 1^er janvier 2026 à vingt-cinq demandes et à vingt-trois décisions, soit plus du double du précédent maximum annuel[5].

La décision n° 2026-341 L du 15 octobre 2026 s’inscrit dans ce mouvement, mais son objet lui confère une portée particulière. Saisi le 15 septembre 2026 par le Premier ministre, le Conseil devait se prononcer sur des dispositions réparties entre vingt-trois articles du livre VII du code de l’éducation, relatives à la gouvernance des universités et de leurs composantes, aux règles communes aux établissements publics à caractère scientifique, culturel et professionnel, à la discipline des enseignants-chercheurs et aux adaptations propres à l’outre-mer. La saisine intervenait huit jours après le dépôt du rapport de la commission d’enquête sénatoriale consacrée à l’excellence académique des universités. Ce rapport recommande notamment de réduire le nombre des membres du conseil d’administration, d’y faire siéger au moins pour moitié des personnalités extérieures et de renforcer la liberté des universités d’adapter leur organisation[6]. Statuant dans le délai d’un mois que lui impartit l’article 25 de l’ordonnance du 7 novembre 1958, le Conseil a déclaré réglementaires toutes ces dispositions, à la seule exception de celles qui organisent la récusation d’un membre de la section disciplinaire compétente à l’égard des enseignants-chercheurs.

Il lui revenait de déterminer si ces règles relèvent des règles constitutives d’une catégorie d’établissements publics ou des garanties fondamentales que l’article 34 de la Constitution réserve à la loi, ou si elles se réduisent à des modalités d’organisation interne que le décret peut modifier. La décision intéresse au premier chef le contentieux administratif, puisque les garanties que la loi cesse de porter relèveront désormais du contrôle du Conseil d’État. Si le Conseil constitutionnel réduit les règles constitutives de la catégorie à ses organes de direction et d’administration, en transposant au contentieux du déclassement une définition forgée contre l’incompétence négative (I), il partage les garanties constitutionnelles entre une réserve de loi limitée à l’impartialité juridictionnelle et le contrôle du Conseil d’État (II).

I. La réduction des règles constitutives de la catégorie aux organes de direction et d’administration

Le Conseil ne redéfinit pas les règles constitutives des établissements publics à caractère scientifique, culturel et professionnel. Il transpose au contentieux du déclassement la définition qu’il avait forgée en 1993 pour censurer l’incompétence négative du législateur (A), puis en déduit une qualification extensive des modalités d’organisation interne, qui gagne les conditions d’élection et de désignation ainsi que le rôle des organes (B).

A. La transposition au contentieux du déclassement d’une définition forgée contre l’incompétence négative

Le Conseil range parmi les règles constitutives de ces établissements « la détermination et le rôle de leurs organes de direction et d’administration, les conditions de leur élection ou de leur désignation, la détermination des catégories de personnes représentées au sein des conseils des établissements, celle des catégories de ressources dont peuvent bénéficier ces établissements, la nature et les fonctions des composantes internes ainsi que les conditions de désignation ou d’élection de leurs organes de direction et d’administration, dès lors que ces composantes sont dotées de compétences qui leur sont propres[7] ». Cette énumération reproduit littéralement le considérant 6 de la décision du 28 juillet 1993, jusqu’à l’ordre inhabituel des adjectifs qui désigne des établissements « à caractère culturel, scientifique et professionnel ». La définition servait pourtant alors la compétence du législateur. La loi déférée autorisait les établissements à déroger par leurs statuts à la plupart des règles d’organisation fixées par la loi du 26 janvier 1984, notamment à celles relatives à l’élection des conseils, à la place des professeurs parmi les représentants des enseignants-chercheurs, aux personnalités extérieures et au vote des budgets des composantes. Le Conseil avait jugé que le législateur avait ainsi méconnu sa compétence et n’avait pas assorti de garanties légales la liberté et l’indépendance des enseignants-chercheurs[8]. Chacune de ces matières relève désormais, au moins pour partie, du règlement[9].

La définition n’en procédait pas moins d’une notion née du contentieux du déclassement. Dès 1964, le Conseil jugeait que le législateur fixe les « règles de création » d’une catégorie d’établissements publics, « lesquelles comprennent nécessairement ses règles constitutives », pour en déduire le caractère réglementaire des dispositions qui « ne présentent pas le caractère de règles constitutives[10] ». En 2026, la définition de 1993 retrouve cette fonction de partage et devient la grille au travers de laquelle chaque fragment est examiné. Chaque motivation décrit la disposition par le verbe « se borner », véritable opérateur du déclassement[11], avant de conclure qu’elle ne figure pas au nombre des règles constitutives et ne met pas en cause les autres principes ou règles placés par la Constitution dans le domaine de la loi. Le pivot de ce raisonnement réside dans une catégorie que la décision de 1993 ignorait, celle des dispositions qui ont « pour seul objet de préciser certaines modalités d’organisation interne[12] ».

L’écart avec 1993 tient d’abord à l’objet de l’examen. Le Conseil appréciait alors une faculté globale de dérogation à des articles entiers, qui atteignait nécessairement les règles constitutives qu’ils contenaient, alors qu’il statue désormais sur des fragments isolés par le Gouvernement, souvent réduits à un alinéa ou à quelques mots[13]. Cette méthode suppose que les catégories de personnes représentées s’entendent des grands ensembles que la loi fait siéger dans les conseils, et non de leurs subdivisions. Elle permet de déclasser tant la présence obligatoire de certains profils parmi les personnalités extérieures du conseil d’administration que l’énumération des secteurs dont elles sont issues[14]. La saisine a toutefois laissé dans la loi les représentants des collectivités territoriales et des organismes de recherche[15], de sorte que des sous-catégories de même nature demeurent de forme législative ou rejoignent le règlement selon le seul périmètre de la saisine.

B. L’extension de la qualification de modalité d’organisation interne aux conditions d’élection et de désignation et au rôle des organes

La ligne ainsi tracée traverse plusieurs matières, dont la distinction entre règle constitutive et modalité demeure ténue. Les conditions d’élection, que la définition range pourtant parmi les règles constitutives, font l’objet d’une lecture restrictive. Dans la ligne de sa jurisprudence relative aux chambres consulaires[16], le Conseil qualifie de modalités la durée et le renouvellement des mandats des présidents et des directeurs[17]. Il en va de même du mode d’élection des conseils « au scrutin secret par collèges distincts et au suffrage direct », ainsi que des règles de vacance, de candidature, de vote, d’attribution des sièges et de suppléance[18]. Demeurent dans la loi le principe électif, l’élection du président par le conseil d’administration et la règle qui fait expirer son mandat avec celui des représentants élus des personnels. Un décret pourrait donc organiser l’élection des conseils au suffrage indirect, alors que le Conseil avait lié en 1984 l’indépendance des enseignants-chercheurs autres que les professeurs à « la sincérité de leur suffrage[19] ».

Le rôle des organes reçoit une lecture tout aussi étroite. La voix prépondérante du président en cas de partage égal des voix au sein du conseil d’administration et du conseil académique est rangée parmi les modalités[20], comme le pouvoir du conseil d’administration d’arrêter le budget d’une composante en lieu et place de son conseil[21], que la décision de 1993 rangeait pourtant parmi les matières dont elle refusait l’abandon global[22]. Le critère des compétences propres, que le paragraphe 2 réserve aux composantes internes, est en outre étendu aux organes eux-mêmes. Le bureau du président de l’université et le comité de direction des instituts extérieurs, que la loi ne dote d’aucune compétence propre, relèvent ainsi du règlement[23]. Le critère présente un caractère réflexif, puisque la qualification dépend du législateur, seul à décider qu’un organe est doté de compétences propres.

La distinction entre conditions de désignation et modalités de sélection marque le point le plus délicat. L’obligation de choisir le directeur d’un institut extérieur aux universités parmi les personnels qui ont vocation à y enseigner n’est qu’une modalité de sélection « en vue de sa désignation dans les conditions prévues par la loi[24] », alors qu’elle délimite le vivier des candidats éligibles. La même exigence figure en termes identiques à l’article L. 713-9 pour les directeurs des instituts et écoles internes, qui n’a pas été soumis au Conseil sur ce point. La loi impose donc aux uns une condition d’éligibilité devenue réglementaire pour les autres. Quant à la désignation de l’autorité qui exerce une prérogative de l’État, elle relève du règlement selon une ligne constante[25], que le Conseil applique à la nomination des directeurs d’école et d’instituts extérieurs ainsi qu’à l’autorisation des fondations partenariales[26]. La formule retenue, qui vise l’autorité compétente « au sein de l’État », porte sa propre limite, puisqu’un décret pourra substituer une autorité de l’État à une autre, mais non transférer la prérogative à l’établissement.

II. Le partage des garanties constitutionnelles entre la réserve de loi et le contrôle du Conseil d’État

Le second terme du contrôle vérifie que les dispositions ne mettent pas en cause les autres principes ou règles placés par la Constitution dans le domaine de la loi. Il ne maintient dans la réserve de loi que la récusation, rattachée aux garanties fondamentales accordées aux citoyens pour l’exercice des libertés publiques (A). Les dispositions qui traduisaient l’indépendance des enseignants-chercheurs passent au règlement, dont le Conseil d’État devient le seul juge (B).

A. Le maintien de la récusation dans la loi au titre des garanties fondamentales des libertés publiques

Seule la récusation devant la section disciplinaire compétente à l’égard des enseignants-chercheurs est maintenue dans la loi. Le Conseil qualifie cette section de « juridiction administrative spécialisée[27] ». Cette qualification est conforme à la loi du 6 août 2019, qui a conservé le caractère juridictionnel de cette seule section, tandis que la section compétente à l’égard des usagers devenait un organe administratif[28]. La faculté pour l’enseignant poursuivi de demander la récusation d’un membre de la section est « une garantie légale du principe d’impartialité, indissociable de l’exercice de fonctions juridictionnelles », qui découle de l’article 16 de la Déclaration de 1789. Les dispositions qui l’organisent mettent donc en cause les garanties fondamentales accordées aux citoyens pour l’exercice des libertés publiques. Ce titre, préféré à celui des garanties fondamentales des fonctionnaires, couvre aussi les personnels non titulaires que juge la section. La notion de garantie légale, qui mesure d’ordinaire ce que le législateur doit aux exigences constitutionnelles, sert ici à fixer sa compétence.

La loi conserve le motif de la récusation, tiré d’une raison objective de mettre en doute l’impartialité du membre, et la liste des personnes habilitées à la demander. Par les mots « de récusation ou[29] », la garantie s’étend à la demande du président de l’établissement, du recteur et du médiateur académique, et protège l’impartialité de la juridiction elle-même. Les conditions dans lesquelles la récusation est décidée relèvent du décret en Conseil d’État auquel renvoie déjà l’article L. 712-6-2[30]. La décision rendue sur la récusation ne peut être contestée devant le Conseil national de l’enseignement supérieur et de la recherche qu’avec le jugement au fond, de sorte que le Conseil d’État, juge de cassation de cette juridiction, demeure le garant ultime de l’impartialité disciplinaire. La saisine n’a pas visé, en revanche, la faculté de renvoyer l’affaire à la section d’un autre établissement, qui demeure dans la loi alors qu’elle sert le même principe.

Le rattachement de la récusation à la fonction de juger s’accorde avec le fait que la section compétente à l’égard des usagers n’en connaisse que par un texte réglementaire[31]. Cette section n’en prononce pas moins des sanctions qui peuvent aller jusqu’à l’exclusion définitive de tout établissement public d’enseignement supérieur[32]. Or, le Conseil a rangé l’impartialité parmi les principes « indissociables de l’exercice de pouvoirs de sanction par une autorité administrative indépendante[33] ». Reste ouverte la question de savoir si cette exigence imposerait aussi d’inscrire dans la loi la récusation devant une section qui sanctionne sans juger. À défaut, c’est au juge administratif, saisi des sanctions prononcées, qu’il reviendra d’en assurer le respect.

B. La soumission des modalités déclassées au contrôle direct du Conseil d’État

À l’inverse, les dispositions qui traduisaient l’indépendance des enseignants-chercheurs ont été déclassées sans que ce principe soit mentionné. Tel est le cas de l’élection « par collèges distincts », du renvoi au décret de la composition des collèges électoraux et de la part réservée aux professeurs et aux personnes habilitées à diriger des recherches au sein de la commission de la recherche[34]. La décision de 1984 avait pourtant déduit de l’indépendance des enseignants-chercheurs, garantie pour les professeurs par un principe fondamental reconnu par les lois de la République, l’exigence d’une « représentation propre et authentique » de chacun des deux ensembles, et censuré pour ce motif le collège électoral unique[35]. Elle rattachait cependant la composition des conseils à « l’organisation d’un service public » et jugeait qu’elle ne touchait pas « à la liberté de l’enseignement[36] ». Le Conseil tient désormais, sans s’en expliquer, ces règles pour étrangères aux garanties fondamentales des libertés publiques comme à celles des fonctionnaires. Quant aux garanties légales exigées en 1993, elles pesaient sur le législateur dans l’exercice de la compétence qu’il tient de l’article 34[37]. L’exigence cesse de s’adresser à la loi dès lors que l’élection par collèges distincts sort de son domaine.

Le principe d’indépendance s’impose alors directement au pouvoir réglementaire, et sa garde passe au Conseil d’État, qui l’assure de longue date. Il a annulé le décret qui instituait un collège unique pour les professeurs et les sous-directeurs de laboratoire du Muséum national d’histoire naturelle, alors même qu’une habilitation législative autorisait ce décret à déroger à la loi[38]. Il a jugé illégale l’inclusion des personnels scientifiques des bibliothèques dans le collège des professeurs, qui privait ces derniers d’une représentation propre et authentique[39], et a exigé que les travaux des professeurs ne soient évalués que par leurs pairs[40]. Il interprète encore les textes « au regard du principe fondamental reconnu par les lois de la République de l’indépendance des enseignants-chercheurs[41] ». Le déclassement change toutefois les conditions de ce contrôle. Tant que les règles figuraient dans la loi, le décret qui les appliquait était protégé par l’écran législatif, et le Conseil d’État jugeait inopérant le moyen tiré de principes constitutionnels contre des dispositions « adoptées conformément » à la loi de 1984[42]. Désormais, le décret qui modifiera ces modalités sera confronté directement à la Constitution et aux principes fondamentaux reconnus par les lois de la République[43], par la voie du recours pour excès de pouvoir, dont il connaît en premier et dernier ressort[44], de l’exception d’illégalité ou du refus d’abroger. La question prioritaire de constitutionnalité sera en revanche fermée, les dispositions issues de ce décret ne revêtant pas le caractère de dispositions législatives au sens de l’article 61-1 de la Constitution[45].

Le déplacement n’est pas neutre quant à l’intensité de la protection. Le Conseil d’État a jugé que la décision de 1984 n’avait censuré le collège unique « qu’en tant qu’elle portait atteinte à l’indépendance des professeurs d’université[46] », alors que le Conseil constitutionnel exigeait une représentation propre et authentique pour chacun des deux ensembles d’enseignants-chercheurs. Le risque tient surtout à l’addition des retouches désormais possibles par décret. Un décret pourrait allonger le mandat du président de l’université et en lever la limite de renouvellement, élire les conseils au suffrage indirect, réduire la part des professeurs à la commission de la recherche, supprimer le représentant des organisations de salariés au conseil d’administration ou étendre le pouvoir de ce conseil d’arrêter le budget des composantes[47]. Il pourrait même laisser les statuts de chaque établissement fixer ces modalités, alors que la décision de 1993 interdisait au législateur d’autoriser sans les encadrer des dérogations aux règles constitutives. Le Conseil d’État apprécierait chacune de ces mesures au regard du principe d’indépendance, mais l’équilibre d’ensemble de la gouvernance qu’elles redessineraient échapperait au débat parlementaire.

Ce passage au règlement demeure partiel en ce que la saisine ne visait pas le I de l’article L. 712-3, qui fixe la répartition des sièges du conseil d’administration. Cela signifie bien que la loi y réserve toujours aux professeurs la moitié des représentants des enseignants-chercheurs, alors que la règle comparable de la commission de la recherche relève désormais du règlement. Elle y fixe aussi le nombre des membres du conseil et la part des personnalités extérieures, que le rapport sénatorial propose de modifier. Seul un nouveau déclassement permettrait de les réviser par décret. Il appartiendra alors au Conseil d’État de vérifier que le décret ne touche qu’aux dispositions déclassées, toute autre modification étant entachée d’incompétence au regard de l’article 37, alinéa 2. Le décret restera par ailleurs soumis à des exigences de procédure que la décision tait, puisque la modification des adaptations propres à la Polynésie française et à la Nouvelle-Calédonie[48] demeure subordonnée à la consultation de leurs gouvernements, dont le Conseil d’État sanctionnerait l’omission[49]. Le Parlement peut encore légiférer en ces matières, une disposition réglementaire contenue dans une loi n’étant pas inconstitutionnelle[50], mais le Gouvernement pourra opposer aux amendements qui les toucheraient l’irrecevabilité de l’article 41, dont on a relevé la « désuétude de fait[51] ». La décision lui ouvre ainsi une voie réglementaire qui lui permettra de réviser les modalités de la gouvernance universitaire sans débat parlementaire, et dont l’usage dira si le déclassement fut une opération de simplification ou l’instrument d’une réforme.

[1] Mathieu B., « La part de la loi, la part du règlement. De la limitation de la compétence réglementaire à la limitation de la compétence législative », Pouvoirs 2005, n° 114, p. 73.

[2] Cons. const., 30 juill. 1982, n° 82-143 DC, Loi sur les prix et les revenus, consid. 11.

[3] Avril P., Le Conseil constitutionnel est-il toujours le bras armé du Gouvernement dans le parlementarisme rationalisé ? Nouv. Cah. Cons. const. 2016, n° 50, p. 39.

[4] Herzog G, « Déclasser la loi pour gouverner : une possibilité limitée pour le futur Gouvernement », Questions constitutionnelles, 25 juill. 2024.

 [5]Décompte établi le 20 octobre 2026 dans le fonds de jurisprudence constitutionnelle de Légifrance, selon la date des décisions. Le Conseil a rendu 314 décisions de déclassement entre 1959 et 2025, dont 25 entre 1990 et 1999 et 63 entre 2010 et 2019. Le précédent maximum annuel était de dix décisions, en 2015. Les demandes n° 2026-321 L à 2026-323 L ont été jointes par une décision unique du 21 mai 2026.

[6]Rapp. Sénat n° 943 (2025-2026), 7 sept. 2026, Ouzoulias P. (prés.) et Garnier L. (rapp.), Universités : assumer l’excellence, recommandation n° 9.

[7]§ 2.

[8]Cons. const., 28 juill. 1993, n° 93-322 DC, Loi relative aux établissements publics à caractère scientifique, culturel et professionnel, consid. 2 à 7 et 12.

[9]§ 5, 7, 24, 26 et 27.

[10]Cons. const., 17 mars 1964, n° 64-27 L, consid. 2 et 4 ; v. aussi Cons. const., 12 déc. 1967, n° 67-47 L, et Cons. const., 6 oct. 1976, n° 76-93 L.

[11]§ 4, 12, 16, 17, 21, 23, 34 et 38.

[12]§ 8 et 29. La formule n’apparaît dans aucune autre décision du fonds de jurisprudence constitutionnelle de Légifrance (recherche en expression exacte du 20 oct. 2026).

[13]P. Avril, art. préc.

[14]§ 5 et 26.

[15]C. éduc., art. L. 712-3, II, 1° et 2°.

[16]Cons. const., 2 oct. 2025, n° 2025-314 L.

[17]§ 4, 14, 15 et 20.

[18]§ 24 et 29.

[19]Cons. const., 20 janv. 1984, n° 83-165 DC, Loi relative à l’enseignement supérieur, consid. 26.

[20]§ 5 et 6.

[21]§ 27 et 29.

[22]Décision n° 93-322 DC préc., consid. 5.

[23]§ 4 et 20.

[24]§ 22 et 23.

[25]V. not. Cons. const., 22 nov. 2012, n° 2012-236 L, à propos de l’École des hautes études en santé publique.

[26]§ 15, 16, 20, 21, 28 et 29.

[27]§ 11.

[28]L. n° 2019-828, 6 août 2019, de transformation de la fonction publique, art. 33, modifiant not. C. éduc., art. L. 232-2, L. 712-6-2 et L. 811-5.

[29]Le dispositif vise ces mots en tant qu’ils figurent « à la dernière phrase du même article », alors que la saisine et le § 10 visent la dernière phrase du troisième alinéa de l’art. L. 712-6-2, la dernière phrase de l’article ne les contenant pas.

[30]C. éduc., art. R. 712-26-1, dans sa rédaction issue du D. n° 2026-36, 29 janv. 2026.

[31]C. éduc., art. R. 811-22.

[32]C. éduc., art. R. 811-13-1, 7°.

[33]Cons. const., 12 oct. 2012, n° 2012-280 QPC, Société Groupe Canal Plus et autre, consid. 16 et 19.

[34]§ 7, 24 et 25.

[35]Décision n° 83-165 DC préc., consid. 20, 27 et 28.

[36]Ibid., consid. 17.

[37]Décision n° 93-322 DC préc., consid. 7 et 12.

[38]CE, 4e et 1re ss-sect. réunies, 29 mai 1992, Association amicale des professeurs titulaires du Muséum national d’histoire naturelle, Lebon.

[39] CE, 4e et 1re ss-sect. réunies, 9 juill. 1997, Picard, Lebon.

[40] CE, 4e et 6e ss-sect. réunies, 22 mars 2000, n° 195638, Lebon.

[41] CE, 4e et 1re ch. réunies, 13 avr. 2018, n° 406887, mentionné aux tables du Lebon.

[42] CE, 4e et 1re ss-sect. réunies, 29 juill. 1994, F. et autres, mentionné aux tables du Lebon ; sur la théorie de la loi-écran, CE, sect., 6 nov. 1936, Arrighi, Lebon, p. 966.

[43] CE, ass., 12 févr. 1960, Société Eky, Lebon, p. 101 ; CE, ass., 3 juill. 1996, n° 169219, Koné, Lebon, p. 255.

[44] CJA, art. R. 311-1, 1°.

[45] Cons. const., 22 juill. 2011, n° 2011-152 QPC, M. Claude C., consid. 4, à propos d’une disposition de forme législative modifiée par décret. Tant qu’elles n’ont pas été modifiées, les dispositions déclassées conservent leur forme législative, et la recevabilité d’une question prioritaire de constitutionnalité dirigée contre elles n’a pas, à notre connaissance, été tranchée.

[46] CE 29 juill. 1994, préc.

[47]§ 4, 5, 7, 24 et 27.

[48]§ 35 à 38.

 [49]L. org. n° 2004-192, 27 févr. 2004, portant statut d’autonomie de la Polynésie française, art. 10 ; L. org. n° 99-209, 19 mars 1999, relative à la Nouvelle-Calédonie, art. 133.

[50] Décision n° 82-143 DC préc., consid. 11.

 [51]Auby J.-B., « L’avenir de la jurisprudence Blocage des prix et des revenus », Nouv. Cah. Cons. const. 2006, n° 19.

These successes are only possible with the cooperation of courageous and tenacious survivors of atrocity crimes. Access to, and support of, those brave individuals willing to share their stories of horrendous treatment are often only possible through the tireless efforts of civil society organizations. These organizations play a crucial role in UJ/ETJ prosecutions and improve accountability efforts in innumerable ways, including by making referrals to and sharing potential evidence with prosecuting authorities, initiating legal proceedings, and representing victims during criminal trials. 

Despite the laudable rise of UJ/ETJ cases, there are simply not enough of them. Widespread calls to expand these prosecutions abound, as do suggestions on how to do so. They include demands from academics, civil society leaders, and even the International Committee of the Red Cross for countries hosting UJ/ETJ prosecutions to offer more political support for these prosecutions, set up new war crimes units, increase the training and resources of existing units, and strengthen national legislation to enhance the use of universal jurisdiction. 

Undoubtedly, those ideas would improve the effectiveness of seeking justice through UJ/ETJ cases, and they are worthwhile recommendations. And certainly, no one doubts that the responsibility to seek accountability for atrocity crimes lies with the States. But realistically, those changes – if they are successful – will be measured in months, if not years. State action can be notoriously slow, and if the increase in prosecutions that the larger human rights community seeks is left entirely to the action of national governments, those working for accountability may be waiting a very long time. But there are things that civil society organizations can do now to increase the likelihood of more UJ/ETJ prosecutions.

How Thinking Like a Prosecutor Can Help CSOs Provide a Boost Now

Eurojust’s Genocide Network and PacificJust, a newer organization extending those efforts to the Pacific Rim, host meetings for personnel of national authorities investigating core international crimes to exchange experiences and best practices. Over my six years at the U.S. Department of Justice overseeing human rights prosecutions, I spent many weeks at such conferences with fellow war crimes prosecutors and law enforcement agents. On these occasions, we had dozens of conversations celebrating investigative wins, brainstorming evidentiary leads, and bemoaning our collective inability to bring more prosecutions. While we complimented the exceptional work and dedication of civil society organizations that brought us evidentiary packages, one area of concern that often came up was how the material provided all too often was insufficient to justify the opening of a formal investigation. 

Recently, I had the opportunity to hear the perspective of one human rights leader on this issue. Antonia Mulvey, Founder and Executive Director of Legal Action Worldwide (LAW), an independent civil society organization that helps survivors of human rights violations pursue justice, acknowledged, “Most civil society submissions have a detailed summary of the alleged conduct and usually the identity of either the victim or perpetrator (or both) involved in an atrocity crime. But often, there’s not much else.” She explained that at LAW their experience has taught them that “a stronger submission is not necessarily a longer one. Instead, stronger submissions require thinking from the outset about what a prosecutor will actually need to act. At LAW, we have increasingly built that thinking into our work. Civil society can make a prosecutor’s decision easier by presenting not simply an allegation, but a realistic route from the evidence already available to an investigation, and potentially to a courtroom.” This is exactly the evolution that will help boost the number of investigations national authorities will consider opening. 

Articulating a roadmap to a successful case is critical. Before a prosecutor can commit the personnel and resources for a full-scale investigation, she must see a route to a viable criminal case. In the United States, for example, a prosecutor should bring charges only when she believes it is more likely than not that the admissible evidence will be sufficient to obtain and sustain a conviction. Allegations of egregious human rights violations abound, and not all of them have the possibility of becoming criminal cases. A case needs an identified and chargeable perpetrator, victims and witnesses willing to testify at trial, evidence corroborating the atrocity alleged, and enough information to defeat the defenses the accused will raise. If an investigation’s prospects are not apparent at the beginning, national authorities will, and should, focus their limited resources on allegations more likely to produce justice for victims.

This isn’t a demand for what does not exist. Since beginning to work with human rights organizations this year following my resignation from the Department of Justice last fall, I have seen how much evidence civil society organizations hold. Often, though certainly not always, such evidence could support a prosecution. But the organization holding it may not recognize that, because recognizing it often takes prosecutorial or courtroom experience. 

While the international community waits for States to strengthen legislation and increase war-crimes-specific prosecutorial bodies, civil society can immediately boost the number of UJ/ETJ cases opened around the globe by giving national authorities fuller evidentiary packages, built the way a prosecutor thinks. 

Better Submissions Result in More Open Cases

How can that be done? In addition to summarizing the evidence of the atrocity crime at issue, civil society organizations should aim to include corroboration of the offenses of the perpetrator or account of the victim already in their possession. They should also identify the investigative leads they can see and explain why each is likely to be fruitful. To be clear – no one expects civil society to perform the work of law enforcement authorities. Most national authorities, for important reasons, prefer that they not. For example, civil society organizations should not take detailed accounts from survivors and witnesses, which could both hamper future prosecutions and harm vulnerable and traumatized individuals. Nor should they conduct multiple interviews of survivors or witnesses to learn more specifics regarding the events. However, by analyzing information already in the organization’s possession with a prosecutorial lens, submissions could be greatly enhanced. 

The existence of corroboration is essential to the success of any prosecution. Civil society organizations should inform law enforcement agents and prosecutors from day one whether there are any eyewitnesses or earwitnesses to the atrocities, medical records reflecting victims’ injuries, photographs or videos related to the events or aftermath, or satellite imagery corroborating the layout of the location as described in witnesses’ accounts. Furthermore, disclose any statements made by the alleged perpetrator about the crimes to others or online; reliable open-source materials that can confirm the affiliation and/or rank of the proposed defendant; or witnesses that can confirm nicknames used by the perpetrator. 

Most importantly – and most often missing from civil society’s evidentiary packages – is evidence or information to rebut the most common defenses asserted by atrocity perpetrators: duress, fabrication, mistake of identity, or collateral damage. When reviewing leads from any source, war crimes prosecutors are not just evaluating whether they can prove that the crime occurred. More often than not, that is the easy part. The most challenging consideration is whether the prosecutor can combat the inevitable defenses. Helping national authorities understand why the obvious defenses do not withstand scrutiny is a crucial element of an opening decision that too often is left unanswered in civil society organizations’ submissions. 

For example, in the face of a fabrication defense, have witnesses with similar accounts lived in separate locations unable to compare stories? If the perpetrator will likely claim duress (and is in a forum where it is available), was the threat to him immediate, and did he have a way out? Can the defendant credibly claim that the entire allegation has been fabricated, and, if so, does the victim have scarring or medical evidence confirming that the physical abuse occurred? For crimes committed during military operations, is there reason to believe that online chatter or overheard orders exist showing that the perpetrators intended to target the victims?

Civil society organizations should have a prosecutorial mindset – or work with others that do – when evaluating their collection of materials. Certainly, some civil society organizations already excel at this task. But there may be even more potential value to the information in the possession of many civil society organizations than they realize. The more thorough, specific, and detailed the evidentiary packages that civil society compiles, the better opportunity they give national authorities to advocate to their superiors that the matter is worth the office’s limited resources to open and begin investigating. 

In Support of Righteous Prosecutors

The world has unquestionably changed over the last 18 months. Justice systems have taken a hit. One analysis estimates that USDOJ has lost over 6,000 years’ worth of agency expertise due to firings or resignations as a result of politicization and the overall numbers of staff have plummeted. The rule of law is facing escalating challenges around the world, and civil liberties groups warn of erosion or regression of the rule of law in several European countries. According to Human Rights Watch, “the global human rights system is in peril.” Against this backdrop, it is not hard to understand why civil society may harbor reservations about collaborating with some national authorities. 

“The answer is not to stop sharing evidence,” Mulvey said while we discussed how organizations can navigate this new environment, “but to be deliberate about who receives it. Civil society can judge an authority on things they can observe: whether its war crime unit is still staffed, whether it has protected witnesses in the past, whether it has kept faith with the organizations that referred cases to it, and whether it has actually brought prosecutions.” I couldn’t agree with her more.

The way forward is following the time-tested path that those seeking justice have long pursued: Work with trusted entities and people, engage familiar faces with a strong track record, and share materials with committed prosecutors and law enforcement agents around the globe whose dedication to the cause of justice can be measured in the results they have obtained on behalf of the most vulnerable victims. They are out there. I have worked with them on joint investigative teams, collaborated with their offices to secure indictments, watched them present torture victims’ accounts at trial, and heard firsthand of the obstacles they overcame to secure convictions. Despite the significant setbacks justice departments have suffered of late, many of the most tenacious, dedicated civil servants I have ever encountered remain, and they are eager to receive evidentiary leads that will justify opening human rights investigations. 

* * *

Universal and extraterritorial jurisdiction cases have been a key element of accountability efforts around the globe, and expanding State capacity and will to pursue such cases, while important, will also take time. In the meantime and in parallel, if civil society organizations are looking to what they can do tomorrow, one solution is clear: By providing national authorities with more specific, corroborated, and detailed evidence – evaluated through a prosecutorial lens – civil society will empower national authorities to increase their investigations and prosecutions of atrocity crimes on behalf of victims and survivors.

The post How Civil Society Can Help Unlock More Universal and Extraterritorial Jurisdiction Cases appeared first on Just Security.

The recent push within the United States to extract hard minerals from the seabed found expression in an April 2025 Executive Order calling for “immediate action to accelerate the responsible development of seabed mineral resources” in three geographic areas: the U.S. seabed (or juridical continental shelf), seabed areas under the national jurisdiction of partner countries, and seabed areas beyond national jurisdiction. The order also provided a non-exhaustive list of strategic minerals “which are vital to our national security and economic prosperity,” including nickel, cobalt, copper, manganese, titanium, and rare earth elements (REE or REY to include the non-lanthanides scandium and yttrium).

Knowing which minerals are the target of commercial seabed mining, in what deposit types they are found, and where, among those three geographic areas, mining activities might take place should assist in understanding which legal regime governs the resource, who the relevant actors might be, and why, how, and when exploration or commercial recovery might proceed. 

What Minerals, in Which Deposit Types? 

The 2025 United States Geological Survey (USGS) List of Critical Minerals contains 60 entries from aluminum and antimony to zinc and zirconium. Of those 60 critical minerals, 43 occur within the marine environment of the U.S. continental shelf, and all 60 occur on U.S. land territory, but most currently are not mined or processed within the United States.

Seabed critical minerals are found in at least six deposit types, each with its predominant mineral composition and unique geographic distribution.

• Polymetallic nodules (PMN) – Small, potato-shaped rocks on the ocean floor at depths of 4,000 to 7,000 meters, containing manganese, cobalt, copper, nickel, titanium, and trace REEs, with concentrations found in the Pacific and Indian Oceans.
• Polymetallic sulfides (or seafloor massive sulfides) (PMS) – Metalliferous accumulations around seafloor hydrothermal vents along spreading ridges at depths between 100 and 7,000 meters containing copper, zinc, lead, and silver.
• Cobalt-rich ferromanganese crusts (CFC) – Accumulations on hard-rock substrates of volcanic origin on seamounts and ridges occurring at depths of 400 to 5,000 meters composed of cobalt, manganese, nickel, platinum, and trace REEs. 
• Rare earth element-rich muds (REE or REY) – Sedimentary layers embedded in the ocean floor subsoil (found at depths below the seabed of 0 to >100 meters) containing concentrations of REEs with known occurrences in the Pacific and Indian Oceans at water depths of 3,000 to 6,000 meters.
• Phosphorites – Sedimentary rocks occurring along continental shelves and seamounts at depths of <1,000 meters containing phosphate, uranium, and trace REEs.
• Marine placers – Heavy mineral sands deposited in nearshore marine environments of <200 meters depth containing titanium, tin, platinum, silver, and trace REEs.

Although often lumped together, literally in the case of nodules, seabed critical minerals satisfy different end uses. REEs are inputs for heat-resistant, permanent magnets; copper is needed for electrical wiring; and nickel, manganese, and cobalt are ingredients in the eponymous NMC batteries. The reasons for their criticality and their processing requirements also vary widely.

Where Matters

The legal rules applicable to the exploration and exploitation of these seabed mineral deposits depend upon where they are found: within U.S. jurisdiction, within the jurisdiction of another coastal State, or beyond the jurisdiction of any coastal State. 

Seabed Within U.S. Jurisdiction

The U.S. continental shelf consists of 11 million square kilometers of seabed extending up to and beyond 200 nautical miles from U.S. territory, including the continental United States, Alaska, Hawaii, and, importantly in the context of seabed mining, the Pacific island territories of American Samoa, Guam, Northern Mariana Islands, and U.S. Minor Outlying Islands. One million of these 11 million square kilometers – the so-called extended continental shelf – lies beyond 200 nautical miles. In this vast area of ocean floor, the United States enjoys exclusive jurisdiction over and sovereign rights to the mineral resources on the seabed and in the subsoil. 

All six deposit types are known to exist within the U.S. seabed area. The Outer Continental Shelf Lands Act, among other U.S. statutes, governs the use of the non-living natural resources of the U.S. continental shelf. The Bureau of Ocean Energy Management (BOEM, now under the newly-formed Marine Minerals Administration at Department of the Interior) is the administrative and leasing agency for these resources. The National Oceanic and Atmospheric Administration (NOAA, under Department of Commerce), which plays a role in all three geographic areas, is the primary marine scientific research agency on the U.S. continental shelf.

Except where domestic seabed mining activities might interfere with high seas freedoms enjoyed by other nations, in particular their freedoms of navigation and laying of submarine cables and pipelines, or where environmental harms from such activities might spread beyond the U.S. continental shelf, public international law has little to say about the modes by which these U.S. resources may be utilized. Instead, in these areas, decisions related to mining will be subject to domestic political and legal processes. 

To offer some concrete examples from current events, community and conservation groups brought a legal challenge in U.S. District Court to block a forthcoming BOEM seabed mineral lease sale for areas surrounding American Samoa, and an unsolicited request to lease phosphorites and heavy mineral sands off the coast of Virginia triggered a BOEM request for information and interest eliciting nearly 77,000 public comments as of July 2026. 

Seabed Within the Jurisdiction of Other Coastal States

Some 150 other coastal States enjoy the same exclusive jurisdiction and sovereign rights with respect to their own continental shelves and seabed natural resources. They, too, control whether, when, where, how, and by whom those resources will be studied or developed. Pursuant to their national laws and under the administration of their government agencies, several of these States are pursuing mineral resource development in their respective seabed areas.

U.S. seabed mining companies, often through subsidiaries, anticipate or are already operating in some of those foreign jurisdictions. At the intergovernmental level, the United States has entered into non-binding agreements (e.g., with Japan and Cook Islands) to expand cooperation and access in respect of areas and resources within the exclusive national jurisdiction of other States. The United States has no jurisdiction or sovereign rights in these seabed areas, but U.S. agencies can play a role here; in partnership with coastal States, NOAA conducts marine scientific research within their zones.

In these exclusive national areas, if and how seabed mining moves forward will depend on the political and legal environments in each individual country. For example, in December 2025, following a national election, Norway, on the verge of opening its inaugural seabed minerals licensing round, instead announced a pause through 2029. In 2023, perhaps in response to the 2019 failure of its Solwara 1 PMS mining project, Papua New Guinea (along with four other Pacific island nations) signed a declaration calling for a region-wide moratorium on seabed mining. In 2022, the Cook Islands issued three five-year exploration licenses, but the prospect of exploitation is not without domestic opposition. In February 2026, Japan collected REY-rich muds from its seabed and plans to return for a mining trial in 2027. 

Seabed Beyond National Jurisdiction

The legal landscape changes substantially in areas beyond national jurisdiction. Approximately half of the planet’s ocean floor is covered by seabed within the exclusive national jurisdiction of coastal States. The remaining half lying beyond, or seaward of, coastal State jurisdiction is the third, legally-distinct part of the seabed referenced in the April 2025 Executive Order.

The rules governing this seabed area (“the Area” (UNCLOS, Art. 1)) and its resources (“all solid, liquid or gaseous mineral resources” (UNCLOS, Art. 133(a))) are set out in the United Nations Convention on the Law of the Sea (UNCLOS or the Convention). The Convention designates the Area and its resources as the “common heritage of mankind” (UNCLOS, Art. 136) and establishes the International Seabed Authority (ISA) to organize, carry out, and control the exploration and exploitation of the Area and its resources on behalf of humankind as a whole. To that end, the ISA has developed exploration regulations for PMN, PMS, and CFC deposits. The ISA has not developed regulations for REY-rich muds or phosphorites, but they would also come under the ISA’s purview. (Marine placers are not found in the Area.)

To date, the ISA has entered into 32 exploration contracts with 22 separate contractors, each of whom is formally sponsored by a State party to the Convention. The Convention, ISA regulations, and individual contracts govern the exploration activities of ISA contractors in respect of the Area’s resources. The ISA continues to work toward the completion of its exploitation regulations and has not yet entered into any exploitation contracts.

One hundred seventy-one States and the European Union are parties to the Convention and are bound by its rules. The United States, uniquely among industrialized nations, is not one of them. In April 2025, the United States announced its intention to unilaterally issue “seabed mineral exploration licenses and commercial recovery permits in areas beyond national jurisdiction” under its 1980 Deep Seabed Hard Mineral Resources Act (DSHMRA, for which NOAA is the permitting agency). Industry responded. Four days later, The Metals Company USA applied for two DSHMRA exploration licenses and one commercial recovery permit, the first license applications since 1984 and the first DSHMRA permit application ever. As of this writing, six applications for exploration licenses, and one for a commercial recovery permit, have been made public. More are in the pipeline. 

These developments in U.S. practice trigger at least two questions of international law:

Q1: Would the United States violate its international legal obligations by authorizing or undertaking unilateral seabed mining?

A1: Since the United States is not a party to the Convention, as a matter of treaty law it is not bound by the Convention’s rules, but it may be bound by similar rules as a matter of customary international law. Some authors conclude that the United States would breach its customary international law obligations by undertaking unilateral seabed mining in the Area (here and here), while others argue that it would not (here). This question is not likely to be tested before an international court or tribunal because the United States has not consented to their jurisdiction.

Q2: What are the obligations of States parties to the Convention in the context of possible unilateral mining by a non-State party? 

A2: States parties to the Convention are under several relevant treaty obligations: the obligation to ensure that a long list of actors do not participate in unilateral mining activities in the Area, including their nationals and any companies and individuals their nationals effectively control (UNCLOS, Art. 139(1)); the obligation not to recognize any acquisition, claim, or exercise of right in respect of minerals recovered from the Area except in accordance with the Convention (UNCLOS, Art. 137(3)); and anti-alienation (transfer of ownership) obligations related to minerals recovered from the Area (UNCLOS, Art. 137(2)). Several authors have explored the contours of these treaty obligations in more depth (here, here, here, here, here, and here), obligations which, if implemented, would isolate and disable U.S. activities in the Area. These questions are likely to be tested before international courts and tribunals. Should unilateral mining continue to move forward, it could spur a request for an advisory opinion or a contentious case for breach of one of the obligations set out above. 

In the Area, if and how seabed mining unfolds will depend on the multifaceted political and legal environments in:

1. Kingston, Jamaica (seat of the ISA and locus of decision-making related to regulations and contractor discipline, among other issues); 
2. The capitals of 171 States parties (sponsorship decisions, effective implementation of international legal obligations, and decisions related to international adjudication will occur here); 
3. Hamburg, Germany (home of the International Tribunal for the Law of the Sea (ITLOS or the Tribunal) and its Seabed Disputes Chamber which would rule on disputes among States parties concerning activities in the Area); and 
4. Washington, DC (where a new administration might revert to a cooperative stance vis-a-vis the ISA and multilateral ocean governance or stay the current course).

Why, Who, How, and When 

Per the 2025 Executive Order, the urgent need to mine the seabed is driven by a scarcity narrative pitting perceptions of need against stability of supply told through the lens of China’s access to land and seabed mineral resources, dominance in critical minerals processing, and willingness to use critical mineral export restrictions as a geopolitical tool. Shifting demand and supply patterns could reduce or increase pressure on seabed resources. Consider the move away from NMC battery chemistries, development of U.S. land-based sources, and potential gains from improved recycling. In stark contrast to the message coming from the White House, 46 States have called for a pause or moratorium on seabed mining in the Area.

Who might be involved in seabed mining depends in large part on the geographic area in which mining activities might occur. Within areas of national jurisdiction, government agencies, industry, and domestic civil society will be involved, as will the national courts when disputes arise. 

In areas beyond national jurisdiction, the constellation of actors grows. Mining under the Convention involves the ISA and its 171 member States, specific sponsoring States, contractors (including their parents, partners, and auxiliary service providers), and international civil society. Disputes arising under the Convention in respect of “activities in the Area” would come before the ITLOS Seabed Disputes Chamber, while broader disputes related to seabed mining might be heard by the full Tribunal or other international courts.

Now, as a major non-State party (the United States) enters the Area, it brings with it additional actors. The ISA, sole regulator for the last 30 years, faces a potential competitor in NOAA. Industry players engaging in regulatory arbitrage are attempting to secure interests under both the ISA and U.S. regimes. And multiple fora are seized of matters related to unilateral activities in the Area, with active cases before both the Seabed Disputes Chamber and U.S. courts.

How seabed mining might occur at commercial scale involves the mechanical engineering required to extract and lift minerals from the seabed and the economic engineering needed to attract investment to an untested industry in an uncertain political, legal, and financial landscape. Without significant government intervention, including commodity price floors, offtake agreements, grants, and loan guarantees, the commercial viability of seabed mining is in question. Some have characterized deep-sea mining as “a multi-billion-dollar solution to problems that do not exist.”

Exploration of seabed mineral resources has been underway for decades. When exploitation might begin will, again, depend largely on the jurisdiction in which it would occur and the regime under which it would be authorized. In the United States, The Metals Company anticipates that NOAA will issue a commercial recovery permit to mine areas beyond U.S. jurisdiction by the end of 2027, but legal action could push that date further into the future. Likewise, lawsuits could delay the BOEM lease sales set for November (American Samoa) and December (Northern Mariana Islands) 2026.

* * *

Access to critical minerals is not a new concern (see, e.g., the 1939 U.S. Strategic and Critical Materials Stock Piling Act). In the middle of the last century, as U.S. land-based sources of hydrocarbons were diminished by global war, the seabed played a major role in satisfying demand. Will the seabed be the next frontier for hard mineral ores, who will make the rules for their exploitation, and who will benefit from their use?

The post United States Seabed Mineral Activities: Understanding the Legal Landscape appeared first on Just Security.

On September 29, leaders from six companies (Google, Anthropic, OpenAI, Meta, xAI, and Nvidia) signed the White House Accord on Super Intelligence, a set of non-legally binding commitments, which includes a provision that the companies would “[p]artner with an independent external auditor or evaluator to carry out independent assessments of whether the controls, monitoring, and detection are operating as intended.”

Anthropic shared an announcement on October 15 with respect to its commitment to undergo independent evaluation. The company stated it had already started to afford “employee-like” access to METR. Anthropic noted that it was “also working with SecureBio and Accenture as evaluators,” and it was in talks with others, including Hugging Face. The SpaceX CEO replied to Anthropic’s announcement stating his company “will probably do the same.”

Anthropic’s October 15 announcement lists expertise, independence, sufficient access to information, and freedom to publish results as the key components for the external review process. As a first mover in this space, its actions may set the norm. It is worth taking a moment to contemplate whether Anthropic’s path will bring teeth and legitimacy to third-party evaluation of frontier AI or, instead, compromise the goal of meaningful oversight and risk assessment. 

While it is commendable that Anthropic has publicly shared its view of how to operationalize third-party evaluation and has quickly moved forward, a number of questions emerge with respect to this approach to independent oversight. Given the Anthropic CEO’s stated commitment to using the company’s AI leadership position to create a “race to the top” (as well as the company’s statement that it is “learning actively about what kinds of partnership work”), the reflections below are intended to help improve this approach to external evaluation before it crystallizes for the company, becomes the industry norm, or is codified in any potential regulation.

Substantive Safety Standards? What Standards? Whose Standards?

Anthropic’s October 15 announcement skips any discussion of the substantive safety standards that will be used in the audits. Will the safety benchmarks be whatever each company agrees to with a particular third-party auditor? Shouldn’t outside experts (in government, the academic research community, and civil society) be involved in designing them? Shouldn’t the public at least have access to those standards before the audits take place? 

This glaring omission is jarring, especially in light of Politico’s release of Anthropic’s lobbying effort regarding federal AI regulation. The outlet reported that Anthropic’s proposal would allow companies to develop their own safety frameworks and permit “the government to block a model’s release or use only if it poses a foreseeable, imminent risk of more than $1 billion in damages, or of killing or seriously injuring more than 50 people. The risk must stem from the creation or release of a biological weapon, the model evading its developer’s controls or a cyberattack.” To say the least, such safety thresholds are profoundly insufficient to meet the moment, if not altogether shocking to most people. 

Will those be the types of safety thresholds the independent auditors are expected to use in their risk assessments? Will the evaluators themselves have any say in what the controls, monitoring, and detection systems should be to avert certain safety risks, or, as indicated in the Accord on Super Intelligence, will they only look at “whether the controls, monitoring, and detection are operating as intended” by the company? Determining solely whether the company is doing what it intended to do could set too low a threshold or even incentivize companies to lower their own standards in order to perform better in the audits.

In addition, will the evaluators be considering solely safety risks within the U.S. context or, as these are global products, the risks arising abroad as well? If so, do the evaluators have the expertise to consider impacts beyond the United States? 

Surely there would be broad consensus that the basic safety standards need to be substantively appropriate and subject to public scrutiny. Without adequate government regulation (or even industry consensus on those substantive standards), we are left guessing what each bilateral contract between a frontier AI company and an evaluator will say. The lack of regulation means a variety of unknown substantive standards will be applied without public input. It will be challenging for the public to accept the legitimacy of the evaluations with such foundational aspects of the audits enveloped in obscurity at their launch.

Issues of Independence

Former SEC Chair Arthur Levitt once stated “The accounting profession must be like Caesar’s wife. To be suspected is almost as bad as to be convicted.” This observation referred to Caesar divorcing one of his wives (Pompeia) on the mere suspicion of misbehavior because, given his position, the appearance of impropriety was untenable. Given the high stakes for these frontier AI evaluations, similarly high standards will no doubt be used in considering independence issues, which is fundamental to the legitimacy of third-party auditors. 

Anthropic has highlighted the thorny issues that are triggered for independence when it comes to funding these audits. The company has previously proposed that funding should be pooled or government-subsidized, but noted that (given neither option exists) it will engage with some evaluators who have their own funding and pay others directly. 

Anthropic’s October 15 update briefly explains the principle of independence as evaluators “shouldn’t hold equity in the company they’re evaluating, or be financially or operationally dependent on it.” It is true that the actual embedded team members of the evaluator, those working on the audit within the evaluator’s offices (but not embedded), and the entire chain of command up to the head of the evaluation entity should not hold equity in the company under review, but what about members of those employees’ households? And are there other conflicts of interest that should be considered? What if the third-party evaluators are recent alumni of the company under evaluation? And what if there are other substantial links between the highest levels at the company doing the evaluating and the company being evaluated? Is transparency a sufficient mitigation in such instances or are more guardrails needed? While it is notable that the evaluators will come from legal entities distinct from Anthropic and for which Anthropic does not select the board of directors, executives, or staff, a variety of conflict-of-interest issues persist that merit addressing.

Indeed, in some sense, any entity engaging in such audits could risk being “financially or operationally dependent” on the company under review if it earns significant revenue from the process or its donated funding is affected by its performance in such audits. There can be problematic incentives to produce sufficiently pleasing audits to continue funding flows. Moreover, if the company under review has other business dealings with the entity conducting the evaluation, concerns can arise about the creation of questionable incentives influencing the audit in order to continue those other dealings. Even the appearance of such incentives could put auditors in the same boat as Caesar’s wife, i.e., at the very least creating the appearance of impropriety, which eventually undermines the public’s view of the legitimacy of the endeavor. 

Some have argued that “randomized selection of qualified auditors combined with finite periods for observation can go a long way toward addressing [such] concerns.” Would such measures be helpful and sufficient in this context? Are rotations feasible given the level of needed expertise, the number of qualified evaluation firms, and the rapidity with which the technology evolves and issues arise in this space? Further thought should be given to how and whether a self-regulatory system of frontier AI audits can ensure meaningful independence.

Access to Information

Anthropic’s October 15 announcement states evaluators should have “[e]nough access to meaningfully evaluate risks. Beyond access to the latest models, evaluators also need visibility into how those models were trained, into how our safeguards function, and more.” Later it states: “Outside a small number of clearly defined categories (e.g., data about customers, attorney-client privileged information, or information about our model architectures), Anthropic employees are encouraged to share information with them just as they would with a fellow employee working on alignment risk.” Then a footnote explains “[i]nformation about architectures presents a high risk of inadvertent leaks, and its relevance to risk assessment is typically low. When it is highly relevant, we will generally seek to share it, though this will require explicit internal approvals.”

Meaningful transparency is also needed about the contractual arrangements between the company under evaluation and the entity evaluating it for the public to fully understand, among other things, the extent of the evaluators’ access to information instead of being given an illustrative listing of exempted information. For example, the phrase “model architecture” can be interpreted in a variety of ways. Perhaps it is defined with more clarity in the contractual relationship between the company and the evaluators? Will evaluator entities auditing the same company have different levels of access to information for different audits? And will the same evaluator entity be given different access to information across the frontier AI industry or will evaluator entities assert minimum baselines of access to information in order to participate in the process? It will be difficult for the public to accept the legitimacy of such audits without greater clarity on the material aspects of access to information in the potentially myriad bilateral legal arrangements that are coming down the pike. 

Transparency of Outcomes

Anthropic’s October 15 announcement states the evaluators will have the “freedom to publish observations and assessments of a model or a process.” No doubt the importance of public findings and reason-giving in these reports is essential to the legitimacy and utility of any oversight of these companies. The announcement explains that, in the context of its particular contractual arrangement with METR, that means 

“METR will publish its analysis of our risk levels, incidents, or practices. Anthropic can’t veto publication, but we can require them to redact information if it falls into predefined categories like the ones above [e.g., data about customers, attorney-client privileged information, or information about our model architectures]. (Anthropic ultimately decides what information does and does not fall into these categories, although METR reserves the right to make certain statements about the level of access and information it received.)” 

Without access to the governing contractual documents and given the brevity of the post, we have only an illustrative list of topics that can justify Anthropic’s redactions and thus cannot assess the breadth of the redaction power. Nor is it clear if METR can publicly share general reflections about whether the redactions were justified. Reporting seems to have noted from a prior version of Anthropic’s October 15 announcement that METR “retains the right to publicly characterize any redactions and state whether it believes they were justified.” Per footnote 2 of the October 15 announcement, the language in question appears to have been updated to “the right to make certain statements about the level of access and information it received,” which covers a different topic than redactions. That said, language later in the post implies METR has “similar” (but not the same) publication rights as other evaluators who are afforded the right to comment on redactions. Overall, the announcement is unclear about whether METR has such rights, though all evaluators should be provided these rights. 

Beyond Anthropic’s right to redact for unknown reasons, are there specific guardrails in place to ensure the company’s pre-publication review of the audit report (to exercise its redaction power) doesn’t otherwise result in improper influence over the content of the report? News reporting indicates the discussions between companies and evaluators on the topic of what gets released to the public has been quite contentious; it is likely to remain so. 

More Inconvenient Questions 

A key set of unanswered questions stems from the unknown process for what happens after an evaluation is completed. If the evaluators give failing grades to the companies, what happens next? Will the companies be bound under the contractual arrangement with the evaluators to change course? Or will the companies merely be subject to judgment in the court of public opinion? Without regulation in place, will the government launch any kind of investigation based on the evaluators’ reports? 

What if the evaluator issues recommendations for future safety improvements? Must the frontier AI companies at least answer the recommendations publicly or will they just be buried in the flurry of daily frontier AI news? Will the evaluators be open to receiving inputs and concerns from outside stakeholders (such as researchers, NGOs, etc.) or will the evaluators’ in-house expertise be the limit of what is applied in the evaluation?

The questions go on and on. In the absence of a regulatory framework (or even rigorous and transparent scrutiny by academic researchers and civil society underpinning the oversight of frontier AI safety), many issues about the independence of evaluators, the substantive standards, and other key topics remain murky, complicated, and problematic (to put it mildly) at this critical juncture. Without discounting the effort that it has no doubt taken within Anthropic to get to the current stage in the development of independent evaluation, the need for substantial improvement is evident and time is short. 

These audits should not be about doling out participation trophies to some of the wealthiest and most powerful companies on the planet. They must be about meeting the moment. More needs to be done to create a meaningful, viable, effective, and transparent evaluation system for frontier AI. Otherwise, the audits will end up at best like Caesar’s wife (sidelined and under suspicion) and at worst set up to fail in the critical mission of averting what the frontier AI CEOs themselves have warned are grave risks to humanity. 

The post Frontier AI and Independent Evaluations: Some Inconvenient Questions appeared first on Just Security.

A curated weekday guide to major news and developments over the last 24 hours. Here’s today’s news:

IRAN WAR

Iran’s Islamic Revolutionary Guard Corps Navy today said four vessels involved in illegal oil transfers in the northern Persian Gulf were simultaneously hit and caught fire after being tracked by Iranian forces yesterday. The IRGC added that three more vessels collided with mines west of the Strait of Hormuz, and warned the U.S. military that continued provocations could lead to a wider response against its regional bases. Tala Ramadan reports for Reuters.

The U.S. and Israeli militaries have developed specific plans for renewed strikes against Iran after the U.S. midterms, U.S. officials and other sources told the New York Times. The countries’ militaries have stepped up exercises, flying planes over Iraq to practice strike maneuvers. The strikes would likely include Iran’s energy infrastructure, nuclear facilities, and missile and drone arsenal. Julian E. Barnes, Eric Schmitt, Farnaz Fassihi, Natan Odenheimer, and Ronen Bergman report.

Dozens of Israeli Air Force fighter jets on Monday conducted a sortie, flying as far as the Iranian border, Israeli and U.S. officials told Axios. The operation took place on the same day U.S. Defense Secretary Pete Hegseth met with Israeli Prime Minister Benjamin Netanyahu, but wasn’t connected to Hegseth’s visit. Barak Ravid reports.

IRAN WAR – YEMEN

Saudi Arabia’s General Authority of Civil Aviation yesterday blamed the Iran-aligned Houthis for a new attack on the eastern Saudi city of Dammam’s airport which killed two people and injured another eight. The U.S. State Department said it ordered the departure of nonemergency U.S. government employees and their families from the kingdom in response to increased hostilities against Saudi Arabia by the Houthis. David Gritten reports for BBC News; James T. Areddy and Abeer Ayyoub report for the Wall Street Journal.

IRAN WAR – OTHER DEVELOPMENTS

The Greek Finance Ministry today said it was implementing sanctions linked to Iran and that it backed efforts to effectively combat money laundering and the financing of terrorism. The statement comes after the U.S. Treasury Department yesterday said it imposed sanctions on Karnakis & Karnakis, a law firm based in Greece and Panama, saying the firm had for years provided legal assistance to Iranian shipping networks. The Treasury also said the U.S. Financial Crimes Enforcement Network issued an administrative subpoena to advance an investigation into a Greek bank, which it did not name. Angeliki Koutantou reports for Reuters.

Palantir Chairman Peter Thiel at a private lecture on Oct. 12 said he is skeptical of the Iran war and does not believe the United States should have launched it, according to audio obtained by POLITICO. Thiel said maybe the United States doesn’t have “the hard power. Maybe we’re going to lose the soft power war.” Calder McHugh reports.

WEST BANK DEVELOPMENTS

U.N. Special Rapporteur on Human Rights in the occupied Palestinian Territories Francesca Albanese today told Reuters that Indonesia should reconsider its involvement in President Trump’s “Board of Peace” because it is being shaped by Israel and the United States and lacks a Palestinian voice. Albanese said Indonesia has historically supported the Palestinian cause, but joining the board was dangerous. Indonesia’s Foreign Ministry said the Board of Peace discussions remained on hold, reiterating its earlier position. Stanley Widianto and Gibran Naiyyar Peshimam report.

New York Democrat congressional candidate Brad Lander yesterday said he was briefly detained and released by the Israeli military while he was supporting Palestinian farmers during an olive harvest in the occupied West Bank. The New Israel Fund said approximately 90 volunteers of what they’re calling the Protect the Harvest initiative were released. Two Israeli lawyers told NPR it was the largest mass detention of activists in the occupied West Bank in at least a decade. Israel’s Foreign Ministry said Lander’s actions amounted to “political provocation,” and noted that he chose to enter a restricted area and participate in an unauthorized activity without coordinating with local authorities. Joseph De Avila reports for the Wall Street Journal; Ruth Sherlock, Samantha Balaban, Nuha Musleh, and Daniel Estrin report.

RUSSIA-UKRAINE WAR

U.S. Special Envoy Steve Witkoff yesterday called Russian President Vladimir Putin a “humanitarian” following the release of U.S. citizen Stephen James Hubbard from Russian custody. In a meeting between the pair, Putin told Witkoff that he “agrees in principle” to resuming trilateral peace talks between Russia, Ukraine and the United States to end the war in Ukraine, Russian and U.S. officials told Axios. Ukrainian President Volodymyr Zelenskyy said he hoped the three-way talks will be held as soon as possible, preferably no later than early November. David E. Sanger, Kim Barker, and Ivan Nechepurenko report for the New York Times; Barak Ravid reports; Ron Popeski reports for Reuters.

Ukraine’s Kyiv Mayor Vitaliy Klitschko yesterday said a Russian missile attack was launched at Ukraine’s capital, injuring six people. The attack came hours after Putin said he was “ready in principle” to resume peace talks and after Trump said the world was “the closest we’ve been” to ending the war. Russia said Ukraine launched 630 drones towards its capital Moscow overnight. Jaroslav Lukiv reports for BBC News; Olena Harmash and Gleb Bryanski report for Reuters.

The Director of the NATO Support and Procurement Agency Jacqueline Borsboom told the Financial Times that NATO will allow Ukraine to buy weapons, ammunition, and military equipment via the agency. Borsboom said the partnership with NSPA “provides Ukraine with the ability to have a more structured framework to leverage a broad range of capabilities that we offer.” Barbara Moens reports.

Russia-linked hackers used a decentralized volunteer-run network to conceal communications during an attack on a Ukrainian grain logistics company, marking the first documented case of a state-linked group using the network this way, according to Slovakia-based cybersecurity firm ESET. The attack culminated in the destruction of data at the company in Sep. 2025 after the attackers gained access to the company’s systems in Jul. 2025 through a phishing email carrying a malicious attachment. Leo Marchandon reports for Reuters.

U.S. Senate Intelligence Committee Vice Chair Mark Warner (D-VA) and House Intelligence ranking member Jim Himes (D-CT) yesterday said the Russian government launched a covert online influence campaign to sway public opinion and undermine bipartisan candidates sympathetic to Ukraine ahead of the U.S. midterms. The pair said the findings were detailed in a classified report on foreign election interference sent to their offices by the U.S. intelligence community earlier this week. Warner said the report suggested that Russia is using AI tools to spread the ideas faster and target U.S. citizens more effectively. Maggie Miller reports for POLITICO.

OTHER GLOBAL DEVELOPMENTS

The U.K. House of Commons Defense Committee yesterday warned that the United Kingdom has “glaring gaps” in its ability to defend the Arctic region, leaving it vulnerable to potential Russian attacks. The committee said a shortage of Royal Navy warships and submarines was leaving the country’s critical infrastructure in the High North vulnerable. Helen Sullivan reports for BBC News.

Italy’s Interior Ministry today said it has agreed to lift border controls with Spain on each other’s citizens. Italy first suspended its open-border arrangements with Spain on Jul. 31, saying it wanted to make sure no migrants entering Ceuta could travel on to its territory. Spain responded by taking similar action against Italy, calling the Italian measures “unjust, contrary to European Union interests, and discriminatory to the population of Spain.” Alvise Armellini reports for Reuters.

CLIMATE DEVELOPMENTS

The U.N. World Meteorological Organization yesterday said the El Niño weather pattern could cause a record rise in ocean temperatures in the Pacific as it barrels across the world like a “wrecking ball,” causing floods and droughts. The WMO’s seasonal outlook predicted the El Niño event would become extremely strong and cause temperatures to rise by 3.8 degrees Celsius. WMO Secretary-General Celeste Saulo said “we have not seen in our records 3.8. We have seen around 3 in some record El Niños before, with very substantive impacts across the world.” Olivia Le Poidevin reports for Reuters.

Thousands of Indonesian protestors this week have taken to the streets to voice anger over the government’s failure to stop fires in rainforests. The demonstrations began on Monday and have spread to more than a half dozen cities across the country. The protests demand Indonesian President Prabowo Subianto to declare a national emergency and take steps to prevent destructive agricultural practices that many blame for fueling the fires. Muktita Suhartono reports for the New York Times.

AI DEVELOPMENTS

CEO of global connectivity and internet security company Cloudflare Matthew Prince yesterday said the company saw thousands of attempts to exploit a flaw AI identified 10 hours before global hacks were publicly announced. Developers of the PHP programming language released urgent new fixes to vulnerabilities that had been lurking in their software for years until they were uncovered by AI. Government officials across the world are warning companies to apply a patch to fix the bugs or consider taking down their sites to avoid an attack. Sam Sabin reports for Axios; Raphael Satter, Leo Marchandon, Kanjyik Ghosh, and Rajveer Pardesi report for Reuters.

Company executives are twice as likely as rank-and-file workers to trust AI over their own judgement, according to a survey from Charter published yesterday. Charter in August surveyed 5,002 U.S. office workers at companies with at least 100 employees who said they use AI at work. 32.8% of executives said they trust AI more than their own judgement, while only 16.3% of other workers said the same. Emily Peck reports for Axios.

U.S. FOREIGN AFFAIRS

The Pentagon is looking to build a more robust deep-cover spying program to conduct economic warfare against China and other U.S. competitors, sources told the Washington Post. Under the new initiative, U.S. operatives would go undercover as business people or other private sector personnel without any overt U.S. government affiliation. Zach Dorfman and Warren P. Strobel report.

French President Emmanuel Macron yesterday said that Trump was one among numerous global leaders who had made false claims about the unrest in France to advance their own political interests. Macron said all “the people who are speaking ill of our country right now are people who don’t like our country,” and that leaders “who echo what certain circles obviously want people to believe about what happened in France, they’re campaigning at our expense.” Ségolène Le Stradic reports for the New York Times.

The U.S. Embassy in Ottawa, Canada yesterday accused prominent Canadian politicians of interfering in U.S. elections and distributed a “fact sheet” to journalists. The document said that the Canadian government “brushes off concerns about potential interference in U.S. elections, even as its ministers and premiers openly describe plans to influence American politics.” The unsigned document singles out Canadian Industry Minister Mélanie Joly, Ontario Premier Doug Ford, Liberal MP Anthony Housefather, and Canada’s consul general in Boston, Massachusetts, Bernadette Jordan. Canadian Prime Minister Mark Carney said that Trump’s tariffs have driven up costs for U.S. families and upended a decades-old trade relationship. Nick Taylor-Vaisey reports for POLITICO; Rob Gillies reports for AP News.

U.S. IMMIGRATION DEVELOPMENTS

House Rep. Ilhan Omar (D-MN) yesterday said she will move out of her Washington D.C. residence that federal agents searched last week. Omar said the attention brought on by the raid made her feel unsafe in the house, and in response to the raid said she is “confident in the fact that I didn’t do anything wrong.” Hannah Schoenbaum reports for AP News.

U.S. DOMESTIC DEVELOPMENTS

Florida Attorney General James Uthmeier yesterday said Kissimmee Mayor Jackie Espinosa was arrested on charges of grant theft and organized scheme to defraud in connection with $50,000 in Covid-19 grant funds that she and her family received from a city program Espinosa helped create. Espinosa has been the mayor of Kissimmee since Nov. 2024. Uthmeier said “Florida taxpayers funded that money to help small businesses recover, not to enrich the mayor who controlled it.” Christine Hauser and Alexandra E. Petri report for the New York Times.

New Mexico yesterday filed a civil lawsuit against sex offender Jeffrey Epstein’s longtime lawyer and accountant, Darren Indyke and Richard Kahn, accusing the pair of helping finance and conceal Epstein’s sex trafficking at his remote Zorro Ranch property in Stanley, New Mexico. The suit alleges several Jane Does, including minors, were abused at the ranch, and Indyke arranged payments to victims that left them indebted to Epstein, while Kahn monitored the spending of victims and even prepared tax returns for some. Khadeeja Safdar reports for the Wall Street Journal.

A New Mexico judge yesterday ordered Meta to pay $25 billion in penalties after a jury in September found that the company misled users about their data privacy. The penalties come nearly a decade after the Cambridge Analytica scandal first broke. New Mexico is the only state to take its claims to trial after 48 others settled with Meta. Josephine Walker reports for Axios.

Meta yesterday warned CEO Mark Zuckerberg could leave California if the state enacts a billionaire tax. The company said Zuckerberg had informed its special committee of the possibility, and the committee decided to give him a financial incentive to continue leading the company from California for the next eight years, despite the risk of a “significant, multi-billion dollar tax payment.” Christine Mui reports for POLITICO.

Chairman of the Federal Communications Commission Brendan Carr on social media yesterday said that broadcast networks could violate their regulatory obligations by airing “entirely fake polls and suppression polls.” Carr said there had been a significant surge in polls “designed to do nothing more than interfere with the regular functioning of the electoral process and alter the normal patterns of voter turnout and enthusiasm.” David Shepardson reports for Reuters.

TRUMP ADMINISTRATION ACTION

The Justice Department’s inspector general yesterday declined to investigate the handling of the criminal case against former Olympian David Hearn, who was cleared of accusations that he intentionally damaged the Lincoln Memorial Reflecting Pool, according to a letter to Hearn’s lawyer. The internal watchdog told Hearn’s lawyer that the allegations of professional misconduct they made fall outside of its jurisdiction and are instead within the purview of the department’s Office of Professional Responsibility. Eric Tucker reports for AP News.

Vice President JD Vance yesterday defended an advertisement featuring the estranged mother of the Democrat nominee for Ohio governor, Dr. Amy Acton. In the ad, Dr. Acton’s mother said her daughter had lied and exaggerated about her childhood struggles. While the ad did not directly mention sexual abuse, it was met with outrage given Dr. Acton’s stepfather was charged with raping the candidate when she was 12. Vance lauded the ad, saying it “calls it out like it is, and calls Amy Acton a liar,” and criticized the response by some Republicans to the campaign ad. Campbell Robertson and Bayliss Wagner report for the New York Times; Ford McCracken reports for The Hill.

White House Counsel Will Scharf in a letter on Monday warned POLITICO that it had 48 hours to retract or correct recent stories about alleged staff drama, or it would suffer unspecified consequences. The letter singled out three stories concerning White House Chief of Staff Susie Wiles despite “on-the-record refutations” or contrary evidence supplied by Trump and senior staff. Marc Caputo reports for Axios.

TRUMP ADMINISTRATION LITIGATION

The Supreme Court yesterday refused to allow the Trump administration to immediately enforce new medical restrictions for transgender people in federal prison. The brief leaves a lower-court block on the policy in place, and does not detail its legal reasoning. Justices Samuel Alito and Clarence Thomas would have allowed the policy to be enforced as the lawsuit played out in lower courts. Lindsay Whitehurst reports for AP News.

A federal judge yesterday blocked the Department of Homeland Security from carrying out most field operations focused on uncovering voter fraud in nine states. The ruling covered stops and field interviews being carried out by hundreds of DHS agents assigned to fan out across the country in search of instances of voter fraud, under what the Trump administration has referred to as the “Unlawful Voter Initiative.” Zach Montague and Shawn Hubler report for the New York Times.

Solicitor General D. John Sauer yesterday submitted an emergency application to the Supreme Court to allow the Trump administration to move forward with construction of a border wall along hundreds of miles in the Big Bend region of southwest Texas. A federal judge blocked construction of the barriers which had drawn local objections. Sauer called the lower court “baseless” and asserted that halting the section of the wall at issue would leave “a 300-plus-mile gap in border defenses.” Abbie VanSickle reports for the New York Times.

Democratic groups on Tuesday sued the Trump administration, accusing it of using unfounded voter fraud claims to influence how states run elections. The suit said the administration is issuing a stream of threats to intimidate and bully both voters and election officials, and condemns the targeting of state voter rolls by insisting that local officials review federally compiled lists of alleged noncitizens on their voter rolls. Josephine Walker reports for Axios.

Did you miss this? Stay up-to-date with our Litigation Tracker: Legal Challenges to Trump Administration Actions.

If you enjoy listening, Just Security’s analytic articles are also available in audio form on the justsecurity.org website.

ICYMI: Yesterday on Just Security

Just the Docs: DHS Discloses July Order Stripping DeepSeek AI from Federal Civilian IT Systems

By Matthew Nagle

The post Early Edition: October 22, 2026 appeared first on Just Security.

Related Just Security coverage: Test, Standardize, Restrict: A U.S. Policy for Chinese AI Models; What DeepSeek Really Changes About AI Competition; Hiding in Plain Sight: The Geopolitics of Software Supply Chains

You can read the document here and it is provided in full below:

The post Just the Docs: DHS Discloses July Order Stripping DeepSeek AI from Federal Civilian IT Systems appeared first on Just Security.

Dr. Terry Kupers said he concluded that the mental illness of defendant Abd al-Rahim al-Nashiri, 61, coupled with brain damage that is visible on an MRI scan that was taken at Guantanamo in 2017, make him unable to competently assist in his own defense and understand the court proceedings against him.

Al-Nashiri, a Saudi citizen, is accused of having a key role in the suicide bombing of the U.S. Navy warship off Aden, Yemen on Oct. 12, 2000. Seventeen U.S. sailors were killed in the al-Qaeda attack, and dozens more were wounded. Jury selection in the case is now scheduled to start Monday, Nov. 2.

Kupers was testifying this week, via video from California, as the first of up to five medical witnesses as part of an eve-of-trial competency hearing ordered by the judge, Army Col. Matthew Fitzgerald.

Al-Nashiri’s lawyers recently presented the judge with two expert opinions, including Kupers’, that declare the defendant mentally incompetent. In response, the judge ordered a military evaluation called a Sanity Board to be conducted. The Army sent a psychiatrist, Maj. Anthony Ralston, who found al-Nashiri fit for trial. Neither opinion has been made public.

Each psychiatrist spent a week at the base, talking to the prisoner and reading medical charts and other classified documents to draw their contrary conclusions. 

Al-Nashiri’s mind is so scattered, Kupers said, that the doctor quickly abandoned the idea of administering a baseline, 20-question test used by mental health experts to determine whether a defendant understands enough of what is going on in court to assist in his own defense.

“He cannot track what’s going on in the courtroom,” Kupers said.

In 2013, two years after al-Nashiri was first charged, a military medical panel diagnosed al-Nashiri with Post Traumatic Stress Disorder and depression but declared him competent to face trial. He has been attending pretrial hearings since then, sitting passively in court with a headset offering translation of the proceedings or watching on a screen in a holding cell near the court. Many hearings put a focus on his four years in CIA custody, 2002-06.

In 2023, an earlier judge suppressed his confessions at Guantanamo as tainted by his torture, which included beatings, waterboarding, being kept naked, and threatened by revving a drill near his hooded head. Kupers said he diagnosed al-Nashiri with severe, chronic PTSD, multiple traumatic brain injuries, a neuro-cognitive disorder, and major depressive disorder.

The prisoner does not believe he is ill, Kupers said. But the doctor said five hours of conversations with the assistance of a trusted translator, as well as an examination of medical records and classified material, showed al-Nashiri suffers from memory loss and confusion that at times prevent him from linear thought and collaborative conversations.

Kupers listed a series of behaviors that underpinned his diagnoses:

• Al-Nashiri needs the lights left on in his solitary cell around the clock because he cannot sleep in the dark, a condition the doctor linked to a period when CIA interrogators kept him in a dark, cramped wooden kennel-like box and would rattle and pound on it from time to time.
• He has cut holes in a privacy curtain that guards installed near the toilet in his cell because he finds sitting behind it claustrophobic. 
• He cannot follow his lawyers’ explanations of ongoing proceedings. When consulted on defense strategy, including whether he would plead guilty to avoid a death-penalty trial, he consistently replies “You decide.”
• He is prone to sudden disruptive outbursts when in the company of other former CIA prisoners on his cellblock. 
• He has dementia-like deficits, does not recognize guards who should be familiar to him and has forgotten the names of lawyers with whom he has spent long stretches, and previously greeted in court with effusive hugs and handshakes.
• Heather Holmes, a psychologist who consulted on the case for the defense, said her first meeting with the prisoner was cordial and then in the next meeting, he yelled, cursed, and ranted for two hours. His hands trembled as he shouted his distrust of his lawyers and the guards. 

A prosecutor questioned Kupers on whether the prisoner might be malingering, or tailoring his answers to the psychiatrist’s questions to elude justice. Kupers said the symptoms demonstrate his diagnoses. On a question from Allison Miller, the prisoner’s capital defender, Kupers said al-Nashiri does not acknowledge his mental disorders or deficits.

The two psychiatrists–Kupers and Ralston–visited on different weeks. 

Kupers testified that he observed the prisoner during a pretrial proceeding from the back of the court and spent nearly six hours with him in three separate meetings in an oversized holding cell near the courtroom. He also visited Camp 5, a maximum-security prison where all 15 prisoners are held at Guantanamo, and al-Nashiri showed him around.

Ralson visited during a week when there were no court hearings and so questioned al-Nashiri at the prison, not at the courthouse.

For his 10- to 15-minute ride to court, prison doctors medicate al-Nashiri with an anti-nausea drug to prevent him from vomiting on himself, and then give him a stimulant meant to treat Attention Deficit Hyperactivity Disorder, ADHD, to perk him up to participate in the proceedings.

Prosecutors have described the cocktail as addressing motion sickness. Defense lawyers have said that traveling in the windowless security van gives the prisoner flashbacks to periods of his torture, and one doctor suggested he is suffering vertigo related to a CIA “enhanced interrogation technique” of ramming a prisoner’s head into a wall to disorient him. The judge rejected a request from al-Nashiri’s lawyers to let him spend trial nights at the court complex, in an oversized, wheelchair-accessible holding cell.

This week’s testimony also put a spotlight on one of the longer-hidden practices the CIA used in its overseas network of black sites: rectal feeding.

Earlier medical testimony showed it was done to al-Nashiri in 2004. To carry it out, agency personnel used a tube that is designed to be put in a patient’s windpipe and instead, inserted it into al-Nashiri’s anus. Using a syringe, they injected a protein-enriched nutritional shake into his body.

The topic triggered sharp exchanges in court between the prosecutor and psychiatrist over whether it was torture or a sound medical response to hunger striking.

Navy Cmdr. Clayton Trivett, the lead prosecutor in the Cole case, told Kupers that two former presidents, James Garfield in 1881 and William McKinnley in 1901, were rectally rehydrated after they were both shot by assassins.

Kupers replied that one can be hydrated through a rectum but not nourished. He said the use of it on al-Nashiri served no medical purpose and was done without the patient’s consent. He called using the technique sadism in the guise of responding to a hunger strike and torture in the form of rape.

Trivett asked if the doctor believed that CIA staff carried it out “for the sexual gratification of those who did it.”

Kupers responded that rape is about power, not sex.

The post Guantanamo Dispatches: Ahead of Jury Selection, Defense Psychiatrist Calls USS Cole Defendant Unfit for Trial appeared first on Just Security.

Background 

The requirement for agencies to publish AI use-case inventories began with an executive order signed by President Trump in 2020. It was later codified into law in the Advancing American AI Act in 2022, which requires each agency to prepare, maintain, and publicly post an inventory of its AI use cases. President Biden’s Office of Management and Budget (OMB) issued a memo in 2024 that directed federal agencies to implement certain safeguards when using AI in ways that could significantly affect people’s rights or safety and required agencies to disclose how they had implemented those safeguards in their inventories. In April 2025, OMB rescinded and replaced the Biden-era memo on federal use of AI. While the 2025 OMB memo has many similarities to the 2024 version, it cut some crucial protections, such as the requirement that agencies measure and mitigate bias. More recently, DHS issued a directive on the use and acquisition of artificial intelligence that permits the chief AI officer to waive any departmental policy governing AI, including any prohibitions on uses; while the directive adds that waivers may not violate “law, regulation, and government-wide policy,” this authority creates an enormous loophole in the existing guidance. 

The inventories that DHS released during the Biden administration had their own transparency problems. Moreover, many of the concerning surveillance tools that Customs and Border Protection (CBP) and U.S. Immigration and Customs Enforcement (ICE) are using today have been in place for years. However, the costs of inadequate transparency are higher now, under an administration that seeks to arrest thousands of migrants a day while undermining efforts to hold federal agents accountable for abuses of power.

How DHS Avoids Required Safeguards

Under the 2025 OMB memo, agencies must report whether a given use of AI is “high-impact” in light of its potential effects on individuals’ rights or safety. The memo requires agencies to implement and disclose certain minimum safeguards for all high-impact AI use cases, including testing the system’s accuracy before deployment, auditing for misuse, and establishing a process to appeal decisions made using AI.

An agency’s use of an AI tool is “high-impact” only if its outputs are “a principal basis” for an official action that affects someone’s rights or safety. As we have written previously, this standard, which the Biden-era memo also relied upon, is far too narrow. Agencies often consider many factors when making decisions, such as a determination by ICE or CBP agents to investigate or detain someone. If an AI’s recommendations, predictions, or other outputs influence such a consequential decision, then agents’ use of that AI should be subject to safeguards, whether or not the AI is a “principal” basis for the decision.

Applying OMB’s narrow standard, DHS avoids implementing safeguards for many uses of AI that directly affect people’s rights. For example, the inventory does not classify the ELITE app as high-impact, although ICE agents use it to find and arrest migrants who may be in the country illegally. The app uses large language models to extract addresses where migrants may live from government records, including data from the Department of Health and Human Services, and displays them on maps, along with “confidence scores” indicating the likelihood that the addresses are accurate, facilitating widescale raids. 

DHS also excludes the Investigative Prioritization Aggregator software, which ranks suspects in criminal investigations by Homeland Security Investigations (HSI), a component of ICE, according to their likely importance to an investigation. Detaining or investigating people affects their rights by definition, and the risks are heightened under an administration that is not scrupulously abiding by its legal obligations and has carried out lethal military strikes of people on boats accused of smuggling drugs. Whether or not these AI systems provide a “principal basis” for decisions about whom to arrest or investigate, they directly shape those decisions. 

Similarly, the department determined that its use of Dataminr, a tool that scans social media and other public sources to “provide alerts for possible threats related to national security [and] border violence,” among other issues, was not high-impact, because the alerts just supply CBP officers with compiled information. This overlooks the substantial concerns associated with use of social media data, including the barriers to accurately decoding social media communications and the risk of inaccurately labeling someone a threat based on tainted, misinterpreted, or inadequately vetted data. 

DHS’s labels have consequences. Excluding technologies from the high-impact list allows DHS to skip safeguards—such as testing and auditing—making it easier for CBP or ICE to deploy technologies that are inaccurate or unreliable, increasing the risk of agents arresting innocent people. Wrongful arrest attempts can be deadly, as tragically underscored when, in two separate incidents during a single week in July, ICE agents killed Lorenzo Salgado Araujo and Joan Sebastian Guerrero, after apparently confusing each of them with other individuals. Interactions with CBP officers can be high stakes as well, from searches of travelers’ electronic devices at the border to fatal encounters. Safeguards are also needed to prevent intentional misuse of AI systems, as detailed below.

DHS Falls Short of the Agency’s Own Rules

The 2025 OMB memo instructs agencies to presume that certain uses of AI are high-impact, and rightly includes activities such as identifying criminal suspects and assessing risks posed by individuals. In follow-up guidance released a few months after the memo, OMB further instructs agencies to explain any deviations from this list.

However, DHS has not always done so. For example, although identifying criminal suspects is presumptively high-impact, DHS does not explain why the “AI-Enhanced ICE Tip Processing” system—which summarizes and ranks the urgency of tips submitted to ICE—is not listed as high-impact. This is not a harmless bureaucratic error. Determining which tips should be investigated determines which threats to public safety are prioritized, affecting the rights of individuals deemed to be suspects and potential crime victims alike. 

What the Inventory Doesn’t Say

For high-impact uses of AI, the OMB memo requires agencies to implement certain measures to manage the technologies’ risks. ICE classifies six and CBP classifies 19 actively deployed use cases as high-impact, but the inventory reveals little about what safeguards are in place. OMB’s follow-up instructions on the AI inventories require agencies to answer nine questions about their compliance with the minimum safeguards for high-impact use cases. All but one of those questions are multiple choice, each with a list of three to six possible responses that convey very little.

For example, CBP and ICE have deployed a high-impact facial recognition app called Mobile Fortify, which agents use to identify individuals they stop to question, potentially matching them to records in federal databases. This is indeed high-stakes technology. A false match could result in wrongful arrest or deportation. However, the agencies are not required to tell the public any meaningful details about their efforts to mitigate these risks. In answer to a question about whether they monitor the use of Mobile Fortify for harms to privacy, civil rights, or civil liberties, DHS simply selected a response that says: “Yes, sufficient monitoring protocols have been established.” 

This generic response is essentially useless, particularly in light of DHS’s removal of the only relevant policy from its website, along with reporting suggesting that ICE officials have told agents they can rely on Mobile Fortify results as “definitive” evidence of someone’s migration status and can ignore contrary evidence – even a birth certificate. The administration’s decision to stop requiring agencies to address demographic disparities in AI models compounds this problem, since facial recognition systems are generally less accurate on women and minorities.

In addition, many of the questions that the Biden-era OMB instructions required agencies to answer were dropped from the Trump-era instructions, including: 

• whether the AI use case has been tested in operational environments to understand its performance and potential impact on affected individuals or communities; 
• whether a senior privacy official has assessed the privacy risks; 
• whether the AI is used to significantly influence or inform decisions or actions that could have a negative impact on specific individuals or groups; and 
• what steps the agency has taken to detect and mitigate significant disparities in performance across demographic groups. 

Excluding these kinds of granular questions deprives the public of important insights into how their government is using AI.

Another gap in transparency stems from the fact that agencies are only legally required to release inventories annually, which allows them to deploy technologies in potentially harmful ways for months before the public learns much if anything about them. For instance, Mobile Fortify became operational at the start of May 2025, according to the inventory, but its existence would have remained a secret until January 2026 if 404 Media had not broken the story. As it was, the public did not learn key details, such as the company that developed the app, until the inventory was published; based on that information, WIRED reported that the app was not designed to confirm identities, as ICE agents are reportedly doing. If ICE agents deployed new AI surveillance tools during Operation Metro Surge in Minnesota in January 2026, the public might not learn about it until 2027. 

Similarly, recent reporting indicates that U.S. Citizenship and Immigration Services (USCIS) is seeking to hire a contractor who may use artificial intelligence tools to help sift through mountains of governmental and commercial data as part of an effort to ramp up criminal investigations of fraud in immigration benefit applications. If the contract is awarded this year, an analysis of the impact of the tools and any risk mitigation efforts would be released in January 2027 at the earliest; if it is not awarded until next year, any disclosures are not likely to appear in the inventory until 2028. 

ICE’s Contradictory Use of Sentiment Analysis

The inventory also reveals that ICE is violating its own policies on internet surveillance by employing automated sentiment analysis. An entry titled “Open-Source Intelligence for Investigations” reveals that HSI uses software systems that search online content for undefined “risks” and suggest “investigative leads.” These systems utilize sentiment analysis: AI methods for classifying the emotions or opinions expressed in text. The entry itself cites an official ICE policy that states the agency is prohibited from applying sentiment analysis technologies to online content during law enforcement investigations.

The policy as written is right to ban sentiment analysis as a tool of criminal investigations. There is little reason to think that sentiment analysis systems could reliably identify real threats to public safety or legitimate criminal investigative leads, since AI models often misinterpret the tone or context of informal comments people make online. Even if such models were perfectly accurate, equipping ICE investigators with tools to search social media for ideas they consider threatening is ripe for abuse, particularly in an administration that conflates criticism with threats and regards viewpoints it dislikes as signs of terrorism. Moreover, Trump administration officials openly announced their intentions to use AI to monitor the internet for perceived support for Hamas among foreign students, which has often meant simply expressions of support for Palestinian rights, and ICE has sought to monitor social media for criticism of the agency. If this is the kind of AI usage the administration is willing to tell us about voluntarily, stronger transparency requirements would likely reveal considerably more. 

What Congress Should Do

In the long term, Congress should pass several legislative reforms to rein in dangerous uses of AI. The Advancing American AI Act leaves it to each administration to decide what details agencies must disclose, which the current administration has interpreted narrowly to weaken the requirement. The Act should be amended to require federal agencies to disclose more meaningful detail in their inventories about how they are using AI. That must include reporting on the safeguards that agencies have put in place for any AI system that impacts people’s rights or safety. To limit ambiguity, the law should explicitly list certain applications of AI for which agencies must report safeguards, including AI systems used to inform decisions about whom to investigate, arrest, or deport. 

Agencies should also be required to update their inventories when a new use of AI is rolled out that has a significant impact – or would be presumed to have such an impact – on rights or safety. In the meantime, the House and Senate committees with jurisdiction over DHS ought to exercise their full oversight powers to bring this administration’s use of AI surveillance technologies out of the shadows, including compelling DHS officials to answer questions under oath.

The post DHS’s AI Inventory Reveals the Limits of Federal AI Transparency Law appeared first on Just Security.

Nine years later, in July 2023, one of the two founders of Rusich walked into Helsinki-Vantaa Airport en route to a family vacation in Nice. He was traveling as Voislav Torden, a name he adopted in a legal name change after the European Union sanctioned him in December 2022 under his birth name, Yan Petrovsky, for his connections to the Wagner Group and his role in leading one of the most notoriously brutal Russian paramilitary formations to emerge from the war in Donbas.

He initially entered Finland on July 19, 2023 with his wife and children at the Vaalimaa road border station in southeastern Finland on his wife’s student visa and his dependent permit. A Finnish border guard’s inspection turned up a facial recognition match to Petrovsky, the name he had used when he fought alongside Wagner forces in Ukraine and Syria. The next morning, Finnish Border Guard officers detained him at the airport as he and his family prepared to board the plane bound for France.

Torden’s airport arrest and eventual conviction by a Finnish lower court in March 2025 for war crimes arising from the ambush on the Aidar Battalion became one of Europe’s most important recent tests of whether a domestic court can prosecute a member of Russia’s irregular forces for crimes committed years earlier in another country’s war. His conviction and life sentence marked the first successful use of universal jurisdiction anywhere to convict a member of a Russian mercenary or irregular formation for war crimes committed in Ukraine before Russia’s full-scale invasion in 2022. 

On Aug. 31, the Helsinki Court of Appeal reaffirmed the central legal significance of that judgment, again finding the 2014 fighting to be an international armed conflict and holding Torden criminally responsible for war crimes, personally and as deputy commander of Rusich. Torden’s failure to properly supervise the actions of two subordinates effectively under his command and control resulted in the commission of a war crime. But while the appellate court has now convicted Torden on the count that carries legal responsibility for the 22 Ukrainian soldiers killed in the ambush itself, it acquitted him of the willful killing of four further soldiers from the same column who had been wounded or trapped by the initial strike, and of the mutilation of one more — reducing his life sentence to 12.5 years. That reduction has been widely reported as a partial victory for the defense. Compared with the broader victory for international justice embodied in the appellate ruling, however, that is a modest win for a convicted war criminal who tried and failed to recast himself as a marginal participant in a conflict that the defense disputed as non-international in character. 

Russia’s Irregular Forces

The ruling carries significance beyond this specific case. For a decade, the Kremlin’s gray-zone strategy in Ukraine, Syria, and Africa’s Sahel region has relied on the calculated ambiguity of formations like the Wagner Group, its Africa Corps successor, and affiliates such as Rusich. State-adjacent, contractually detached, deniable in Moscow, and until recently unreachable in Western courts, they operated in the space between Russian defense policy and Russian organized crime. 

Much of what is known about Russia’s irregular forces from 2014 to 2023 stems from leaked digital files from the late Wagner leader Yevgeny Prigozhin’s shell companies, detailing in granular fashion the timeline of his consultations with top-tier Kremlin decisionmakers, and painting a complex picture of Wagner’s internal power structures. Additional colorful details about Russia’s irregular forces can be culled from Wagner’s substantial social media presence. One key detail that emerged from the company files and documents is that many Wagner operatives were issued Russian passports lodged under false names. Research led by one of us (Rondeaux) at New America’s Future Frontlines program, based on leaked Wagner Group records, has catalogued more than 5,800 Wagner personnel who carried false Russian identity documents between 2014 and 2021. The catalogue documents names, call signs, and career trajectories. Both Rusich commanders are referenced in those personnel records. Prosecutors building cases against Wagner-linked personnel can expect, as the Torden case demonstrates, that prominent paramilitary commanders can be traced through open-source documentation.

But they should now also assume the Torden precedent may apply. The appellate court’s landmark decision opens new avenues for potential prosecution of Russian irregular forces, widening the scope of not only who can be brought to account but also answering the question of when Russia’s war in Ukraine began for purposes of international criminal responsibility.

The judgment also recasts the terms of a doctrinal debate about the blurred lines defining proxy warfare that has raged since the International Criminal Tribunal for the former Yugoslavia handed down its judgment in the case of Duško Tadić in 1999. That ruling accepted a lower legal threshold for considering a foreign state’s relationship to an organized armed group sufficient to internationalize a conflict, departing from the stricter “effective control” test the International Court of Justice articulated in Nicaragua v. United States in 1986. The lower test is one of “overall control.”

That question goes to the heart of Russia’s long-running effort to shape the legal terms of proxy warfare by preserving distance between the state and armed groups acting in its interests. Russia’s official story about Ukraine has always been that the war started in February 2022. Everything before that – targeted attacks on civilian infrastructure, torture of prisoners of war, extrajudicial killings of political figures, in Moscow’s telling, was a civil war in which Russia was not a party. The Kremlin asserts that the fighting in Donbas from 2014 was not a clash between two sovereign states, but a non-international armed conflict, a localized struggle between Kyiv and pro-Russian separatists fighting the Ukrainian government. Pro-Russian paramilitary “volunteer units” like Rusich and Wagner Group operating in Donbas before Russia’s full-scale invasion were not, Moscow insists, covert instruments of the Russian state, but independent combatants with no legally meaningful connection to Russian security forces under Putin’s command.

Politically, if the separatist formations were genuinely independent or autonomous, Russia could distance itself from responsibility for their conduct and deny the Kremlin’s military role as peripheral or nonexistent, as Putin often argued until Wagner forces marched on Moscow in open rebellion against the state in June 2023. Putin applied the same obfuscating legalese in labeling the 2022 phase of the war a “special military operation,” while selectively punishing anyone who calls it otherwise and criminalizing the sharing of “false information” about Russia’s armed forces. This obfuscation serves two purposes: domestically, Moscow avoids the general mobilization a declared war would require; internationally, it is Russia’s legal firewall against both state responsibility for the paramilitary formations fighting on its behalf and criminal responsibility for the war crimes those formations commit. In applying its own interpretation of Tadić, the Finnish appellate court exposed a chink in that armor. 

Putin had already weakened his own defense in the summer of 2023. In a nationally televised address three days after Prigozhin abandoned his mutiny against Moscow, Putin admitted on camera that “the financing of the entire Wagner Group was fully ensured by the state,” disclosing that the Russian Ministry of Defense had spent about $1 billion supporting Wagner operations between May 2022 and May 2023. In the language of the Tadić test, he had placed significant evidence of overall control into the public record himself. But the August 2023 plane crash then killed Prigozhin and Dmitry Utkin, Wagner’s top operational commander, removing much of the group’s senior leadership and complicating any effort to establish accountability for Wagner’s conduct in Ukraine, Syria, and parts of Africa.

The circumstances leading to Yan Petrovsky’s decision to legally change his name to Voislav Torden in 2022 are murky. But his personal journey to the heart of the controversies associated with Wagner’s conduct was clearer. Born in the Siberian city of Irkutsk in 1987, he spent his adolescence in the Norwegian coastal town of Tønsberg after his mother remarried. By the spring of 2014, Petrovsky — then still shuttling between Oslo and St. Petersburg — had joined a Russian Imperial Movement paramilitary training program alongside Alexei Milchakov, a self-avowed neo-Nazi; the volunteer detachment they formed there surfaced publicly as Rusich that July, attached to the pro-Russian Batman Battalion fighting in Luhansk. Norwegian police deported Petrovsky to Russia in October 2016 after determining his ties to the right-wing extremist Nordic Resistance Movement and to the Russian ultranationalist underground he had been cultivating from an Oslo tattoo shop. 

Both men trained at the paramilitary program run by the Russian Imperial Movement, which the U.S. State Department designated in April 2020 as the first white-supremacist organization ever listed as a Specially Designated Global Terrorist entity. Rusich fought alongside Wagner Group units in Ukraine in 2014 and 2015, following Wagner to Syria in 2016 and 2017. Torden (aka Petrovsky) is also inextricably linked with Milchakov’s suspected involvement in the brutal murder of Syrian national Hamdi Bouta in 2017. Five years later, Rusich quietly advertised their return to Ukraine in social media postings discovered by New America in January 2022. 

The U.S. Treasury Department sanctioned Petrovsky, Milchakov, and Rusich in September 2022, and the EU followed suit in December 2022. Milchakov commanded the unit until he was wounded near Kharkiv, Ukraine, later that year, at which point Torden (aka Petrovsky) took over. Torden denied his leadership role to Finnish authorities, saying he had been in the Donetsk and Luhansk regions to provide “humanitarian aid” and to engage in “ideological propaganda.” In their later appeal, Torden’s defense team took a slightly different tack, echoing Putin’s rhetoric, contesting the prosecution’s assertion that Russia and Ukraine had been at war in Donbas by September 2014 and that Wagner, Rusich, and the smaller Kremlin-aligned formations were belligerents in it from the outset. (If the war were an internal armed conflict within Ukraine instead of an international one between Ukraine and Russia, it would have narrowed the nature and scope of war crimes prosecutable under Finnish domestic law.)

The Helsinki appellate decision rejected that assertion, and like the lower court aligned with a 2022 ruling by a Dutch court, that found that Russia had exercised overall control over the self-proclaimed Donetsk People’s Republic in eastern Ukraine when pro-Russian separatist forces shot down Malaysia Airlines Flight MH17 killing all 298 people aboard in July 2014. In that case, the Dutch District Court cited Moscow’s provision of weapons, personnel, and training as well as its influence over senior appointments and military operations. The Finnish judiciary’s ruling extended that reasoning to the fighting in Luhansk several months later, providing further judicial recognition that Russia’s war in Ukraine was already international in character years before the February 2022 invasion. 

Four Lessons from the Rusich Case Beyond Helsinki

The decision ensures that, at least outside of Russia, victims of abuses attributed to Russian forces – conventional or irregular – might have a path to remedy, though the modest compensation on offer to the family members of Aidar Battalion fighters raises questions about the fairness of that remedy – a few thousand euros per family member, based on Finnish tort law. 

The Torden prosecution is the first European case brought against a Russian irregular-warfare commander for atrocity crimes committed in Ukraine, and the most substantial European domestic prosecution of any Wagner-adjacent figure to date. A 2024 study led by the War Crimes Research Office at American University identified six European jurisdictions best positioned to prosecute Russian atrocity crimes under universal-jurisdiction principles: Germany, Sweden, France, Finland, the Netherlands, and Norway. The Aug. 31 Finnish ruling is now the reference text for the rest, and it carries four implications.

First, Finland’s evidentiary standards will help define what future prosecutions can realistically achieve. By dismissing the most serious counts where the evidence fell short, the appellate court gave the surviving judgment a degree of credibility that an unaltered life sentence might not have carried. For the broader accountability project, a judgment that can withstand serious scrutiny in a foreign courtroom is more valuable than a maximalist ruling that cannot.

Second, command responsibility on the narrower reading the appellate court applied will make future prosecutions of fighters for Wagner – and possibly Africa Corps – difficult. Prosecutors must show which specific subordinates were under effective command and control of the superior, that those subordinates committed the specific acts alleged, and that the defendant failed to prevent or punish them. The command structures of Russian irregular forces have been documented to a degree. But experience shows that it takes considerable effort to triangulate and move from lead evidence to prosecution.

Third, the case demonstrates the importance of cross-border collaboration. When Finnish investigators needed to identify the voice giving the shooting order on a Rusich cellphone video, they sent the recording to the Estonian Forensic Science Institute. Forensic analysis indicated that the voice recorded did not match Torden’s, which proved decisive in the outcome of the most serious count on appeal. Similarly, mutual legal assistance protocols between Finnish investigators and Ukraine’s Office of the Prosecutor General, and the witness-interview procedures developed for Aidar survivors testifying by video link from Kyiv proved crucial. Interpreters’ work to translate the proceedings will also serve as reusable capacity for the next case in Finland. The model is exportable to the other jurisdictions.

Fourth, the Finnish Border Guard’s detection and subsequent arrest of Torden after his entry into the Schengen visa area highlights the importance of travel bans against Russian fighters as both a law enforcement tool and a critical component of accountability. Torden had previously been barred from entering the Schengen zone under his former identity, Yan Petrovsky, but nonetheless entered Finland under a new name and with a Finnish residence permit. It was a routine border check, intelligence information linking the two identities, and facial-recognition analysis that ultimately exposed the deception and brought him into Finnish custody.

Entry bans and sanctions are often treated primarily as instruments of pressure or exclusion, but the Torden case shows that in practice they can also create the administrative tripwires that place suspected perpetrators within reach of investigators. That logic has gained urgency since Germany’s determination that Russia was behind an attempted August 2026 attack at Leipzig/Halle airport, where an explosive-laden drone was found beside Ukrainian cargo aircraft used in NATO airlift operations. One of the suspected operatives reportedly entered the Schengen area on a tourist visa issued by Italy, prompting renewed EU discussion of tighter Russian visa restrictions and an EU-wide entry ban on current and former Russian combatants.

None of this makes the ruling a straightforward victory for Ukraine. Torden’s sentence reduction and the two dismissed counts will likely be interpreted by Aidar Battalion families as a partial failure to name clearly and with force what happened to their loved ones. Both objections are fair. A Nordic court capable of reversing itself on the evidence in a case that carries this much political weight is nonetheless the kind of court whose convictions will hold when the war ends and the political weather changes. 

Torden’s defense team has publicly stated that it will appeal the decision to the Finnish Supreme Court. Torden’s attorney, Natalia Malgina, told Russia’s state press agency TASS that the aim is full acquittal. As of this writing, there has been no official record of a defense appeal being filed, and no public notice has been provided by the prosecution as to whether it intends to cross-appeal to restore the original counts. Historically, the Supreme Court has rarely granted such leave and only for cases raising novel questions of law. Given the appellate court’s careful record, denial is the likelier outcome. If leave is denied, the appellate ruling becomes the working precedent for European Russian-atrocity prosecutions. For prosecutors building the next file against Russian irregulars in Berlin, The Hague, Paris, Oslo, or Stockholm, it is the closest thing they have to a manual. Milchakov, the Rusich co-founder, was last known to be in Russia according to postings on the unit’s Vkontakte social media account in 2024 (the account has since been taken down. Reportedly Rusich was operating under Russian military cover near the Finnish-Russian border.

The post Finland’s Prosecution of Russian Mercenary: The Rusich Appellate Ruling and War Crimes Accountability appeared first on Just Security.

A curated weekday guide to major news and developments over the last 24 hours. Here’s today’s news:

IRAN WAR

Iranian President Masoud Pezeshkian yesterday said he told Pakistani Army Chf. Field Mar. Asim Munir that there would be no talks with the United States until Iran’s demands were met. According to an Iranian readout of the meeting, Munir also met Iranian parliament speaker Mohammad Baqer Qalibaf, who said the United States could not regain Tehran’s trust until it fulfilled seven conditions Iran set out last month for reopening the Strait of Hormuz. Menna Alaa El-Din and Yasmine Ghania report for Reuters.

IRAN WAR – YEMEN

Yemen’s Iran-aligned Houthis today denied launching any drones towards the Saudi holy city of Medina. The Saudi-led coalition in Yemen earlier said it had intercepted drones over Medina and the wider Eastern region. Nayera Abdallah reports for Reuters.

Iranian Foreign Minister Abbas Araghchi yesterday said Iran has not received a response from Saudi Arabia or the Houthis to its offer to help broker talks to end the fighting between them. Araghchi added that any “foreign intervention [in the conflict] would complicate matters,” and said Egypt has previously told Iran it would “not directly” interfere in the Yemeni conflict. Menna Alaa El-Din and Yasmine Ghania report for Reuters.

A Turkish lawmaker from the ruling AK Party today told Reuters that Turkey will submit to parliament a mandate allowing for military personnel to be deployed to Saudi Arabia under its joint defense pact with the kingdom and Pakistan. Nevzat Devranoglu reports.

IRAN WAR – OTHER DEVELOPMENTS

U.K. counter-terrorism police yesterday said two men have been re-arrested in connection to the incident near Royal Air Force Fairford last month. A U.K. national and dual U.K.-Iranian national are being held on suspicion of assisting a foreign intelligence service under Section 3 of the National Security Act 2023, and the police said the investigation relates to Iran. Cachella Smith reports for BBC News.

Israeli Prime Minister Benjamin Netanyahu is pushing to resume large-scale strikes against Iran and had instructed the Israeli military to plan for a new major offensive earlier this month, sources told the Financial Times. The IDF had prepared for a multi-day bombing campaign to take place on the weekend of Oct. 10 that would have targeted civilian infrastructure, nuclear facilities, and regime targets, and would have surpassed the intensity of the joint U.S.-Israel offensive in Feb. 2026. The sources cited President Trump’s opposition to restarting any conflict before the midterms as why the mission did not go ahead. Neri Zilber reports.

ISRAEL-HAMAS WAR

Israeli strikes and gunfire in two separate incidents in Gaza yesterday killed at least four people, according to local health officials. The IDF later said that Akram Yassin, who was killed alongside his wife, was a senior official in Hamas’ general security apparatus in Gaza which was tasked with protecting senior Hamas officials, but did not comment on the two gunfire incidents, one of which killed a 17-year-old boy. Nidal al-Mughrabi reports for Reuters.

RUSSIA-UKRAINE WAR

Ukrainian police today said four people were killed in overnight Russian attacks that targeted energy facilities across four regions. Two of the dead were in the capital Kyiv, while the other two were killed outside the northeastern city of Kharkiv. Ukrainian Prime Minister Sergii Koretskyi said Russia had attacked power facilities in the Kyiv, Volyn, Odesa, and Mykolaiv regions, as well as civilian enterprises in at least two other regions. Dan Peleschuk reports for Reuters.

Russia’s western Bryansk region Gov. Yegor Kovalchuk today said a Ukrainian drone attack on a car killed a 19-year-old man and injured two people. Reuters reports.

Russia’s Defense Ministry yesterday said the Russian military struck three cargo vessels in Ukraine’s port of Reni, and two other cargo vessels en route to Odesa. Russian forces also hit two logistics hubs in Ukraine’s northwestern Volyn region and in the city of Khmelnitskiy. JekaterĪna Golubkova reports for Reuters.

U.S. Special Envoy Steve Witkoff today met with Russian President Vladimir Putin in St. Petersburg, Russia and presented him with an updated proposal to end the war in Ukraine, a U.S. official told Axios. Witkoff yesterday met with Ukrainian President Volodymyr Zelenskyy and other European representatives in Norway, and following the meeting Zelenskyy said the next step should be a trilateral meeting between himself, Putin, and Trump, a format that Russia has so far rejected. Reuters reports; Barak Ravid reports.

Russia today released U.S. citizen Stephen James Hubbard from Russian custody in an act of goodwill before Putin’s meeting with Witkoff. Trump yesterday said he had asked Putin to release Hubbard. Hubbard was seized from his home in Ukraine months after Russia invaded Ukraine and was accused of being a mercenary. The United States has maintained that the charges against Hubbard were fabricated. Lucy Papachristou reports for Reuters; Luke Broadwater reports for the New York Times.

U.S. Secretary of State Marco Rubio on a phone call yesterday told Zelenskyy that Trump hasn’t issued any kind of order to harm Ukraine. The 45-minute call took place several hours ahead of Zelenskyy’s peace talks with Witkoff. Rubio also acknowledged to Zelenskyy that there could be a delay in sending Patriot interceptors that have already been paid for by NATO countries because the United States needs to make sure it has enough to protect its own military bases around the world. Barak Ravid reports for Axios.

SpaceX CEO Elon Musk last week ordered the slowdown of Starlink, U.S. officials and other sources told the Wall Street Journal. Some sources said Musk capped the speeds at which Ukrainian hardware could stay connected to the satellite-based network, which rendered some offensive and interceptor drones useless, and shortly after the news became public Musk discussed the decision on a call with Trump. Another source said the pair discussed the matter prior to the decision to slow down Starlink. Rubio today appeared to support the suggestion that Musk slowed down Starlink, as he said the decisions regarding Ukraine’s access to Starlink were up to SpaceX and Musk, not the White House. Alexander Ward, Natalie Andrews, and Josh Dawsey report; Gram Slattery reports for Reuters.

Russian Deputy Prime Minister Alexander Novak today said Russia will dispatch its first tanker with diesel for the United States tomorrow. Novak said the vessel will carry 40,000 metric tons of diesel and that the paperwork on the fuel supply is nearing its completion. Vladimir Soldatkin reports for Reuters.

U.S. intelligence agencies say the failed drone attack operation at Germany’s Leipzig-Halle Airport this summer was one of the opening shots in a stepped-up campaign of covert action against NATO allies by Russia, European officials briefed on the findings told the Wall Street Journal. The officials said there was a growing frustration in Russia this summer over the standstill on the ground in the war in Ukraine and Ukraine’s refusal to agree to concessions Putin wanted. Bojan Pancevski reports.

The EU is preparing to try and persuade Canada and the United Kingdom to increase their support for Ukraine as part of an effort to fill a €70 billion funding gap Ukraine faces next year, EU sources told POLITICO. The EU-Canada summit on Oct. 29-30 and the EU-U.K. summit on Nov. 20 will give the EU a chance to make the case. Zoya Sheftalovich and Gerardo Fortuna report.

OTHER GLOBAL DEVELOPMENTS

French Prime Minister Sébastien Lecornu yesterday said the country will reinforce police with troops and mount a huge security operation for upcoming protests over underfunding of education in the country and wider issues planned on Saturday. Lecornu acknowledged the government’s failures in its response to last weekend’s violence and said the deployment “will be massive.” French President Emmanuel Macron today pledged to prevent a repeat of the weekend’s riots. Elizabeth Pineau and Dominique Vidalon report for Reuters; Elizabeth Pineau and Dominique Patton report for Reuters.

Australian lawmaker Helen Haines, who was hit by a car on Monday in what authorities say was a deliberate attack, has lost both her legs, her husband confirmed today. Phil Haines added that “her heart and mind are in good shape” and that the “good news in all this is that she’s alive.” The alleged attacker Brendan Nugent appeared in court today on charges of dangerous driving causing serious injury, intentionally causing serious injury, and causing harm to a Commonwealth public official. Lana Lam and Katy Watson report for BBC News.

Burkina Faso’s Justice Ministry today said their parliament yesterday adopted a revised penal code that restores the death penalty. The revised code reintroduces the death penalty for crimes including terrorism, espionage, treason, and importing expired or banned products that result in death. Reuters reports.

Meta yesterday said it has complied with an order from Brazil’s Superior Electoral Court for the company to give equal treatment on its platforms to President Luiz Inácio Lula da Silva and opposition candidate Sen. Flávio Bolsonaro before Sunday’s presidential runoff. Meta added that the restrictions ordered by the court “do not affect organic posts.” Mauricio Savarese reports for AP News.

CLIMATE DEVELOPMENTS

U.K. Energy Secretary Miatta Fahnbulleh today said “following careful consideration,” the government has made the decision to approve production at Jackdaw gas field off of the east coast of Scotland. Fahnbulleh added that Jackdaw will start supplying gas to homes from as early as the end of the month. Jackdaw was initially approved for production in 2022 but has faced years of litigation and pushback from environmental groups, who insist that more drilling would hamper efforts to cut carbon emissions, and production would not protect energy security. BBC News reports.

The Director of the Potsdam Institute for Climate Impact Research Johan Rockström yesterday told the Reuters Sustainability Europe conference that global warming caused by fossil fuel emissions is accelerating. Rockström said damage to coral reefs and the Amazon rainforest was close to becoming irreversible, and parts of the Atlantic Ocean circulation system were close to collapse. Simon Jessop reports.

A U.N. Environment Program report published today said countries are not doing enough to cut emissions that cause climate change, and that the United States sticks out as the only major country where planet-warming pollution increased dramatically last year. The United States was tied with Russia as the G20 countries with the most per capita emissions of carbon dioxide and other planet-warming gases. The report also warned that global warming is now projected to exceed previous estimates, with even the most optimistic scenario seeing average temperatures rise to 1.7-1.8 degrees Celsius above pre-industrial levels by mid-century. Rebecca Hersher reports for NPR; Simon Jessop reports for Reuters.

AI DEVELOPMENTS

The U.N. Institute for Disarmament Research will launch a new initiative tomorrow on the military uses of AI. Head of UNIDIR Robin Geiss yesterday warned that “the margin for human judgement is increasingly narrow, including in the nuclear domain,” and that it is difficult to assess how close the world is to machines acting outside of human control. Edith M. Lederer reports for AP News.

China yesterday accused the United States of discrimination “under the pretext of so-called national security” after a U.S. order banned products by Chinese AI start-up DeepSeek from the technological systems of many U.S. government agencies. The U.S. Department of Homeland Security’s order, published on Monday but issued in July, cited fears that China may gain access to sensitive U.S. data from DeepSeek to “produce actionable intelligence.” Ephrat Livni reports for the New York Times.

U.S. TECH DEVELOPMENTS

Google yesterday said that several of its cloud computing services remained disrupted around London following an outage. Google said the cause of the “power event” had been dealt with and that it was working to restore service. Leo Sands reports for the New York Times.

U.S. FOREIGN AFFAIRS

U.S. Ambassador to Turkey and Special Envoy for Syria and Iraq Tom Barrack yesterday told Middle East regional security officers at the Global Security Forum in Qatar that “America as the guarantor to the world, for the last 80 years, it is no longer there.” Barrack said what the Trump administration is doing is not an abandonment, but a “realignment,” and that the region is right to pivot toward smaller and more diversified alliances like the new defense pact between Saudi Arabia, Pakistan, and Turkey. Feliz Solomon reports for the Wall Street Journal.

The Liberty Justice Center yesterday filed a lawsuit on behalf of food business Gellert Global Group, urging the Court of International Trade to find the Trump administration’s “liberation day” import taxes against Canada unlawful and to order refunds with interest for affected importers. The suit argues that the White House exceeded its authority in imposing the tariffs and caused “immediate harm to American businesses.” Rachel Clun reports for BBC News.

U.S. IMMIGRATION DEVELOPMENTS

ICE yesterday denied hitting Ricardo Gonzalez with a car while he was cycling in Long Beach, California before detaining him, but video footage obtained by the New York Times contradicts their statement. The footage shows an S.U.V. driving the wrong way down a one-way street and swerving toward an oncoming bicyclist, and an officer then proceeds to jump out and appears to draw a weapon. Long Beach Mayor Rex Richardson (D) submitted a complaint to the DHS Inspector General Office and said Gonzalez has worked for the city. In the video footage, Gonzalez can be heard saying “I’m a U.S. citizen!” Madison Malone Kircher reports.

The State Department yesterday said the Trump administration has revoked over 300,000 visas during Trump’s second term. The number showed an acceleration in the pace of visa revocations in recent weeks, as an Aug. 10 State Department update had put the number of visa revocations at 175,000. Kanishka Singh reports for Reuters.

The FBI yesterday said it is investigating reports of an “incendiary device” being thrown at the ICE building in downtown Atlanta late last night. The FBI said there were no reports of injuries or significant damage. AP News reports.

U.S. DOMESTIC DEVELOPMENTS

A bribery investigation conducted by FBI anti-corruption investigators uncovered by ProPublica showed that before Trump purged the FBI and Justice Department of their teams dedicated to taking on public corruption, the now deceased Sen. Lindsey Graham (R-SC) and other members of Congress were under investigation for taking bribes from a Hawaiian naval defense contractor called Navatek. In early 2025, the FBI opened an investigation into bribery on Capitol Hill and Navatek CEO Martin Kao confessed to federal agents that Navatek’s business model consisted of bribing officials in exchange for government contracts. The FBI took the formal step of naming two former Navatek executives as subjects of a bribery investigation but did not yet designate Graham or others as they were still actively investigating their interactions with Kao. Molly Redden, Avi Asher-Schapiro, William Turton, and Kirsten Berg report.

The FBI released 38 pages of FBI interview notes, surrounding a criminal investigation led by prosecutor John H. Durham over the use of a torture program by the C.I.A. against 9/11 terrorism detainees, to the New York Times in response to a Freedom of Information Act lawsuit. The interview notes were the first tranche to be released, which could offer insights into the controversial program. Charlie Savage reports.

The Mississippi Attorney General’s office on Monday announced a police officer who shot and killed 1-year-old Kohen Wiley while chasing two women accused of shoplifting was justified in shooting at their car in June, and that the state would not file charges as there was “insufficient evidence of criminal conduct.” Although a police officer was not named as the shooter, Senatobia Police Department officer Hunter Foster was the only officer among those present that day who was placed on leave. Body camera footage released yesterday shows the car the two women were in slamming into a parked vehicle, and after it had been ordered to stop, accelerating left and appearing to travel toward at least one of the officers before passing between two of them. At least one officer shoots into the vehicle. Joseph Cranney and Mukta Joshi report for the New York Times; Joseph Cranney, Mukta Joshi, Eduardo Medina, and Devon Lum report for the New York Times.

Supreme Court Justice Elena Kagan yesterday criticized the court’s conservative majority for a series of election-related decisions in recent years that Kagan said have led to more polarization and a dysfunctional democracy. Kagan delivered the critique in front of an ideologically mixed crowd of law students and professors, and said that the court “has kind of blinded itself to the very deleterious effects of their rulings on how the political system works.” Ann E. Marimow reports for the New York Times.

TRUMP ADMINISTRATION ACTION

Trump and Vice President JD Vance yesterday announced the $25 million Fraud Whistleblower Incentive Program which will pay out cash for people who tip off the government about fraudsters. Trump said if “anybody turns anybody in… you get a 1 per cent commission, and I think that will be amazing, because the kind of money that’s been stolen is trillions from this country.” Trump will act as the sole arbitrator of who receives payouts. Mallory Wilson reports for The Hill.

Trump will highlight his administration’s efforts to crack down on foreign criminal networks that lure U.S. citizens into fraudulent crypto investments today, a White House official told Reuters. The official said Homeland Security Secretary Markwayne Mullin, FBI Director Kash Patel, Defense Secretary Pete Hegseth, and Attorney General Todd Blanche are expected to attend the White House event. Courtney Rozen reports.

Federal Emergency Management Agency Administrator Cameron Hamilton yesterday told state officials they must increase their financial reserves and emergency capabilities to handle most disasters on their own, according to a recording of his remarks obtained by POLITICO. Hamilton added that a “future I hope we would shoot for is one where we wouldn’t require as many disasters because the capability of states is so high and so substantial.” Hamilton’s remarks add detail to his previously reported Oct. 15 internal memo. Thomas Frank reports.

TRUMP ADMINISTRATION LITIGATION

CNN is threatening further legal action if its journalists continue to be denied access to Air Force One and the “pool” of journalists who accompany Trump when he travels. Jocelyn Noveck reports for AP News.

Did you miss this? Stay up-to-date with our Litigation Tracker: Legal Challenges to Trump Administration Actions.

If you enjoy listening, Just Security’s analytic articles are also available in audio form on the justsecurity.org website.

ICYMI: Yesterday on Just Security

Foreign Threats to U.S. Elections Are Maturing Faster Than the Defenses to Counter Them

By Brian O’Neill

The post Early Edition: October 21, 2026 appeared first on Just Security.

These actions resemble those documented in previous election cycles, but the capacity to conduct such operations is expanding. Advances in artificial intelligence are lowering technical and financial barriers that once made sophisticated influence operations largely the province of capable states and well-resourced organizations, allowing a wider range of foreign actors to conduct them.

The United States appears capable of detecting and assessing the kinds of foreign activity it has confronted in past elections. It is less certain that the election-security system can adapt at the same pace as foreign capabilities expand, potentially to include attacks on the mechanics of an election itself.

Public reporting to date suggests the United States will get through November despite continued foreign efforts to influence the election. The danger lies in treating such an outcome as evidence that the system is prepared for what follows. November matters not simply as another election, but as the starting point for a race the United States may already be losing.

This does not discount the prospect of interference originating within the United States, including through the exercise of government power, in November or thereafter. Those threats raise a different set of legal and constitutional questions best reserved for a separate discussion. 

What the Public Record Shows

The 2016 presidential election brought foreign interference in U.S. elections to widespread public attention. The threat itself was not new. Russia and other foreign actors had sought to influence American politics before then, but the 2016 election marked a significant escalation in the scale, scope, and methods of foreign interference.

During the 2016 campaign season, Russian actors stole and released political information, ran covert influence operations, and probed state election systems. The Senate Intelligence Committee later found that Russian-linked cyber actors targeted election infrastructure in numerous states and, in a small number of cases, gained access to voter-registration databases. However, its bipartisan review released in 2019 found “no evidence that any votes were changed or that any voting machines were manipulated.” The committee also concluded that Russian activity was “part of a larger campaign to prepare to undermine confidence in the voting process.”

That pattern continued. In September 2018, President Trump issued Executive Order 13848, establishing a formal process for assessing foreign interference after federal elections. The first assessment under that order, following the 2018 midterms, identified influence activities by Russia, China, and Iran, but found no evidence that election infrastructure had been compromised in a way that prevented voting, changed vote counts, or disrupted tallying.

Congress expanded the reporting requirements in 2019, requiring the intelligence community to assess foreign threats and vulnerabilities in state election systems before federal elections and update that assessment closer to Election Day.

In March 2021, the intelligence community concluded that Russia sought to influence the 2020 election through covert influence operations, once again aimed at shaping public perceptions and undermining confidence in the electoral process. The assessment also attributed to Iranian hackers intimidating emails sent to swing-state voters falsely claiming to be from the far-right group the Proud Boys, while China considered but did not deploy influence efforts intended to change the outcome of the presidential election. The assessment found no indications that a foreign actor tried to alter any technical aspect of the voting process, from voter registration to ballot casting, tabulation, and reporting of results.

The 2022 midterms showed the field of actors widening. The intelligence community reported that the scale and scope of foreign activity exceeded what it had detected in 2018 and involved a broader range of actors. Russia, China, and Iran mounted the most significant efforts, while Cuba and several other countries engaged in smaller and more narrowly targeted activity. The assessment also found that foreign actors generally favored influence operations over attempts to interfere with election infrastructure, which were more difficult and carried greater risk.

By 2024, the public warning cycle had accelerated. An October 2024 National Intelligence Council assessment noted Russian and Iranian influence activity during the campaign, including Russian election-related content across text, images, audio, and video and Iranian efforts to denigrate former President Trump. The assessment also warned that China, Iran, and Russia were better prepared to exploit the period after voting ended through information operations and other efforts to sow confusion.

The 2026 reporting so far reflects the same operational focus on shaping perceptions. One notable difference is that the intelligence community’s 2026 Annual Threat Assessment issued in March omitted U.S. election influence and interference entirely. A year earlier, the 2025 assessment explicitly warned that Russia viewed influence operations targeting U.S. elections as advantageous and expected Moscow’s influence activity to increase in sophistication and volume. It also described Chinese and Russian use of AI-enabled influence tools. 

That omission in the 2026 Annual Threat Assessment did not mean collection had stopped. In April 2026, Army Gen. Joshua Rudd, the head of the National Security Agency (NSA) and U.S. Cyber Command, told Congress that foreign adversaries could be expected to try to interfere in the midterms. By September, classified U.S. intelligence assessments reportedly concluded that Moscow had authorized another digital influence campaign, while researchers tied AI-generated fake videos targeting competitive Senate races to a Kremlin-linked network. According to the same reporting, China is focused on individual races where U.S. policy toward Beijing is at issue, while Iran is operating at a smaller scale.

October reporting adds further evidence of such influence operations. The New York Times reports that threat intelligence company Alethea found that a Russian-backed influence operation was exploiting divisions over Israel to target American voters. X separately announced that it had removed 13,000 accounts linked to influence operations connected to Russia, China, and Iran, which amplified content about US politics targeting candidates and public figures across the political spectrum.

One piece of the public record remains missing. The Office of the Director of National Intelligence (ODNI) has not released, as it has in earlier federal election cycles, the formal declassified assessment of foreign activity in the 2024 election. The intelligence material President Trump declassified in July 2026 did not fill that gap; it largely concerned earlier elections and related foreign capabilities and activity. Democratic senators have criticized the failure to release the 2024 assessment — a gap that Senate Intelligence Committee Vice Chair Mark Warner (D-VA) noted in response to Trump’s July release.

The gap in the public record extends beyond the missing 2024 assessment. At his July confirmation hearing, Director of National Intelligence Jay Clayton acknowledged the intelligence community’s responsibility for addressing foreign interference in U.S. elections. Since taking office, however, he has provided little public account of the threat facing the November election or of how the intelligence community is responding to it. On Oct. 2, Sen. Warner, along with Rep. Jim Himes (D-CT), the ranking Democrat on the House Intelligence Committee, wrote Clayton seeking greater transparency about foreign threats to the midterms. They acknowledged that intelligence officials had briefed Congress on the latest threat assessments, but that several statutorily-required products remained outstanding, and called for the responsible release of unclassified assessments before Election Day.

The record of foreign interference over the last decade provides some reassurance, but only about the threats the United States has already learned to recognize — a less reliable guide to what foreign actors may be capable of doing as the tools available to them change. With the U.S. government likely to face more foreign activity, its preparedness for an evolving threat depends in large part on the system the United States spent years after 2016 building to detect, assess, and respond to foreign election interference.

Whether what remains can keep pace with a maturing threat is the issue that will matter most after November.

What Was Built — and What Has Changed

After Russia’s 2016 operation exposed gaps between intelligence agencies, law enforcement, cybersecurity officials, and the states that actually run elections, the federal government spent the next several years building a system intended to connect those pieces. 

The Obama administration in January 2017 designated election infrastructure as critical; the Trump administration created the Cybersecurity and Infrastructure Security Agency (CISA) in 2018 and, through Executive Order 13848, established a recurring process for assessing foreign interference after federal elections. The FBI’s Foreign Influence Task Force and, later, ODNI’s Foreign Malign Influence Center added mechanisms for integrating and responding to foreign influence threats. The center, which Congress authorized in 2019, housed the Election Threats Executive, the intelligence community’s coordinating authority for election security.

The resulting architecture was not a single command structure. Its value lay partly in the relationships connecting intelligence agencies, law enforcement, cybersecurity specialists, state and local election officials, and the private sector. CISA and the federally funded Election Infrastructure Information Sharing and Analysis Center (EI-ISAC), for example, provided election offices with assistance designed not only to detect an operation but also to move information among institutions that held different pieces of the problem.

Much of that connective tissue has changed since 2025. In February 2025, then-Attorney General Pam Bondi disbanded the FBI’s Foreign Influence Task Force as part of a broader shift of Justice Department resources toward cartels and transnational criminal organizations. The administration also curtailed parts of CISA’s election-security work and federal support for the EI-ISAC. CISA lost nearly a third of its staff in early 2025 and has had three acting directors and no confirmed one. Election officials have since reported losing services such as penetration testing, tabletop exercises, and regular threat briefings. 

At the State Department, Secretary Marco Rubio closed the Counter Foreign Information Manipulation and Interference office in April 2025, arguing that its work had become entangled with censorship. Four months later, then-Director of National Intelligence Tulsi Gabbard announced an ODNI restructuring that cut ODNI’s workforce by more than 40 percent. She also directed the dismantling of the Foreign Malign Influence Center as a stand-alone organization, reducing its functions and dispersing the remaining work across other parts of the intelligence community. The Cyber Threat Intelligence Integration Center, which served as a hub for sharing and coordinating classified cyber threat intelligence across the intelligence community and civilian agencies, was also eliminated.

Important capabilities do remain. Press reports suggest that the intelligence community continues to collect and assess foreign election threats, including influence activity. CISA still offers election-security resources, and the NSA and U.S. Cyber Command retain capabilities developed during earlier cycles. Federal funding for state election security fell from $55 million in 2024 to $15 million in 2025, then partially restored to $45 million for 2026; however, state and local election officials continue to describe federal support as inadequate. 

But federal support that once operated on a standing basis has become fragmented and, in some areas, sharply reduced. CISA’s 2026 Election Infrastructure Security Plan arrived only 40 days before the midterms. The joint NSA-Cyber Command Election Security Group—which pools intelligence and cyber operations against foreign election threats, and in 2022 was stood up early in the year. In late September, Defense Secretary Pete Hegseth directed the U.S. Cyber Command and the defense intelligence agencies to prioritize election defense. Officials have not clarified what role, if any, the Election Security Group would play in the activity Hegseth ordered.

The Election Assistance Commission, which certifies voting systems and administers federal election-security grants to states, has had no commissioners since July, when the president fired its two Democratic members and the remaining Republican commissioner resigned. Career staff can continue existing programs, but the commission cannot adopt new policies or update election guidelines without a quorum. State election officials say federal intelligence sharing and technical assistance have become sporadic, while some states have turned to private vendors and their own resources to replace that support.

The interruption of regular federal engagement became more apparent this month. The New York Times reported that intelligence officials attempted to deliver a classified briefing to state election personnel on Oct. 15, but postponed it after a technical problem prevented a secure connection. The session reportedly would have been the first of its kind in more than two years.

The weakness of the federal election-security system is increasingly at the seams between the capabilities that remain. Intelligence officers can still collect, investigators can pursue leads, and states can secure their own systems. But those capabilities are less useful if information is not fused quickly, responsibility is dispersed, or warnings reach election officials too late to act. The post-2016 architecture was built precisely to prevent those pieces from operating in isolation.

The Post-Election Test

The election next month will provide the first opportunity to see how well this diminished system performs under actual election conditions. But even a relatively uneventful midterm would reveal little about whether the United States is prepared for how much the threat could change before 2028. 

The United States will enter the 2028 presidential cycle with a dangerous mismatch: a maturing foreign threat and a diminished system to counter it. States can harden voting systems, conduct audits, use fusion centers, and buy cybersecurity services, but many have room to do more. In April, a Brennan Center survey of 834 local election officials found that 75 percent had received no additional state or local resources to compensate for reductions in federal election-security support. 

Governors and state legislatures should not wait for Washington to restore that support. They can provide local election offices with cybersecurity expertise, strengthen coordination with law enforcement and emergency-management agencies, and press Washington for sustained intelligence sharing and technical assistance. With many new governors taking office in January, there is an opportunity to make election resilience a statewide priority well before the 2028 campaign is underway.

But states cannot see the full dimensions of a foreign campaign that crosses state lines or originates overseas. They cannot replace federal intelligence collection, integrate reporting across the intelligence community, or attribute a foreign operation. Those responsibilities remain with the federal government.

Congress has tools to determine how well the federal system still works and to require action where it does not. The immediate task is to ensure that the intelligence community meets the election-security requirements already in law, including the DNI’s post-election assessment due by mid-December. That assessment, together with the reporting Congress receives before the election, should allow lawmakers to compare what the government knew before November with what actually occurred, what intelligence reached state officials, how quickly foreign activity was attributed, and whether functions once performed by dedicated election-security organizations are still being carried out elsewhere.

That is only the beginning. A Congress of either party must also use hearings, classified oversight, and investigative authority to establish what capabilities have been lost and who is now performing them. Restoring those capabilities is certain to be challenging, especially in the short period before the presidential primaries and given the administration’s approach to election security to date. If the administration remains unwilling to rebuild the system, a divided government could leave the United States entering 2028 with the problem clearly identified but only partly addressed.

Congress also has an immediate obligation to address what may become the greatest challenge a rebuilt election security system will have to confront. Earlier this month, Trump put DNI Jay Clayton in charge of a government-wide AI effort that will recommend how Washington should respond to the technology’s risks and opportunities. The mandate is broad enough to encompass foreign election interference but does not identify it as a specific priority. Congress should insist that it become one. The review should assess how AI is changing foreign election operations and what collection, analytic, attribution, and warning capabilities the government will need before the next presidential election.

Congress has less time to act than the calendar suggests. Rebuilding the election-security system would be difficult even if the administration reversed course tomorrow. Offices can be reestablished and responsibilities reassigned relatively quickly. Experienced analysts and operators cannot be replaced on the same timetable, nor can relationships with state officials, collection expertise, or practiced interagency procedures. 

And the rebuilding would begin at a time when the intelligence community itself has undergone substantial workforce reductions and organizational disruption, while an internal CIA survey, reported this summer, found growing concern among analysts about political pressure on their work. The problem before 2028 is therefore larger than restoring the organizations dismantled since 2025. Rebuilding their capabilities will require expertise, professional independence, and institutional relationships that the government’s broader disruption of intelligence has put at risk.

The post Foreign Threats to U.S. Elections Are Maturing Faster Than the Defenses to Counter Them appeared first on Just Security.

Related Just Security coverage: Mark Nevitt, The New “National Defense Area” at the Southern Border: What You Need to Know (April 29, 2025); Murder by Drone: The Legal and Moral Stakes of the Caribbean Strikes

You can read the documents here and here, and they are provided in full below:

Department of Defense Announcement

Federal Aviation Administration Domestic Notice

The post Just the Docs: Departments of Defense, Transportation, and Homeland Security Announce Southern Border Drone Ban appeared first on Just Security.

Initially, the policy response in the United States occurred entirely at the federal level. But over the past several years, U.S. states have begun to consider legislation on the issue. Not all of this legislation has been adopted, and the U.S. state-level response is clearly in an early phase. As U.S. states continue to respond to transnational repression, it is important that state legislators – and ultimately, members of Congress – understand what the states have done and what can be learned from their various approaches.

One of the challenges of countering transnational repression is that it has not yet been defined in U.S. federal law. The democracy and rights organization Freedom House, which has played a leading role in drawing attention to the issue and in gathering data on it, defines transnational repression as “governments reaching across [national] borders to silence dissent among diaspora and exile communities.” (One of us, Grady, is a research analyst at Freedom House.) Perpetrator governments employ a range of tactics – including assassinations, illegal deportations, abductions, digital threats, and coercion-by-proxy – to intimidate and harass those who pose a perceived threat to the regime. Transnational repression can also involve the abuse by authoritarian regimes of legitimate international legal cooperation mechanisms such as Interpol’s notice system, mutual legal assistance frameworks and counterterrorism financing structures.

The Federal Response

The federal response to transnational repression in the United States has been varied. Numerous congressional committees have held hearings — including multiple hearings by the U.S. Helsinki Commission and most recently, and relevantly, the Congressional-Executive Commission on China (CECC) June 4, 2026, hearing on “The PRC’s Threats to Americans: Transnational Repression & State-Level Responses.” Federal law enforcement has regularly brought cases against perpetrators, and U.S. congressional offices have introduced several bills on transnational repression, including the Transnational Repression Policy Act (TRPA). Yet, that bipartisan and bicameral legislation, which would codify a formal definition of the phenomenon into law and mandate the development of a government strategy to blunt the threat, has struggled to move on the Hill, given other legislative priorities.

U.S. administrations have also evaded the intent of the most significant U.S. legislation, Section 6503 of the 2022 National Defense Authorization Act, entitled “Transnational Repression Accountability and Prevention” or “TRAP.” The provisions directed the Attorney General and Secretary of State to provide biannual reports to Congress on authoritarian governments’ misuse of Interpol’s systems, support reforms to prevent such abuse, and develop a coordinated U.S. response to this form of transnational repression. While the required initial report was submitted in August 2022, the TRAP provisions have not produced the sustained reporting, coordinated implementation, or broader accountability that Congress envisioned. The result is that, while the U.S. federal government has done more than any other government to raise awareness about transnational repression, its response has been neither unified nor consistent.

Indeed, the United States and other democracies continue to facilitate transnational repression via their immigration and extradition systems by detaining and deporting dissidents. Part of the challenge is that, precisely because of the widespread participation of democracies in international legal cooperation, it is all too easy for perpetrator governments to abuse these systems to seek the return of exiled activists and dissidents. While those mechanisms require greater transparency and stronger safeguards against abuse, receiving countries must also ensure that they do not detain, extradite, or deport individuals based solely on uncorroborated allegations or requests from governments with a documented history of abusing these systems for political purposes.

By the same token, a great deal of transnational repression occurs outside of immigration enforcement, in the everyday lives of those targeted. Adhering to due process standards when adjudicating claims for asylum, while essential, will not protect individuals from surveillance, harassment, coercion, or intimidation by foreign regimes or their agents inside the United States. These are acts that can be addressed at the state level. Indeed, given the federal structure of the United States, leaving every offense to the federal government to criminalize is neither sensible nor safe. U.S. states, in addition to serving as legislative laboratories, can play a constructive role in responding to the challenge of transnational repression.

State Legislative History and Legal Mechanisms 

In recent years, state legislatures have increasingly taken steps to address foreign interference. In general, legislative initiatives have been broad, treating transnational repression, alongside land sales and foreign lobbying campaigns, as increasingly serious national security concerns. These broad approaches have prompted some concerns about the potential for overreach. At a June hearing of the U.S House of Representatives Select Committee on China, the president of Asian Americans Advancing Justice, John C. Yang, cautioned that efforts to address foreign interference must avoid repeating historical instances of national security concerns leading to ethnicity-based suspicion rather than evidence-based enforcement.

Since the beginning of 2025, seven states have adopted laws that explicitly mention transnational repression and enact measures to respond to the phenomenon. While these laws vary to some extent in scope and approach, they reflect growing state-level attention to the problem.

In May 2025, Texas became the first state to pass legislation to counter transnational repression. Senate Bill 1349, signed into law by Governor Greg Abbott on May 24, 2025, and effective on September 1, 2025, criminalizes actions committed by a foreign government or terrorist organization, or an agent working on behalf of these entities, when the acts interfere with Texas residents’ protected conduct under the First Amendment of the U.S. Constitution or in some way coerce a Texas resident to either work for a foreign principal or to leave the United States. The law also makes it illegal to enforce another country’s laws in the state. The sentencing enhancement for serving a foreign principal can be applied to pre-existing offenses in the Texas penal code, including assault, stalking, and human trafficking. The law also requires the Texas Department of Public Safety to create a training program for Texas law enforcement on common tactics and on the primary perpetrators of transnational repression, and requires the Department, in collaboration with the Governor’s office, to prepare a study on the issue.

While the law identifies how transnational repression infringes on the rights of affected individuals, it takes an overly expansive view of the perpetrators by including terrorist organizations. Unlike terrorist offenses, which usually attack state institutions or the public at large, government-directed campaigns are targeted against particular private individuals who, as individuals, have the rights the law seeks to protect. Moreover, the act incorrectly implies that terrorist organizations stably control territory and are recognized foreign principals.

Soon after Texas passed its law in 2025, Nebraska adopted the Crush Transnational Repression in Nebraska Act. It defines transnational repression as actions committed by a foreign government or terrorist organization, or an agent working on behalf of these entities, “with the intention of harassing, intimidating, censoring, or otherwise extending such principal’s ability to influence, control, or impose such principal’s preferences on the behavior of individuals outside of the principal’s territory or jurisdiction.” Similarly to the Texas measure, the Nebraska law introduces a criminal sentencing enhancement for covered offenses — including first-, second- and third-degree assault, stalking, and violation of a harassment protection order, among others — that interfere with Nebraskans’ protected conduct under the First Amendment of the U.S. Constitution or that seek to force a Nebraska resident to work for a foreign principal or leave the United States.

In addition to these increased penalties, the act calls on Nebraska State Patrol to create training on cross-border repression tactics and key perpetrators, to establish a website with a multilingual reporting mechanism for targeted individuals, and to prepare a report for the legislature. Certainly, training and multilingual reporting tools are helpful. However, the legislation’s repeated reference to the term “foreign adversary” may increase the likelihood that the implementation process will only focus on governments like China and Iran and overlook transnational repression linked to U.S. allies such as Saudi Arabia and India.

Laws in Arizona (2025), Kansas (2026), Louisiana (2026), and Tennessee (2026) share very similar descriptions of transnational repression and policy prescriptions with the legislation approved in Texas and Nebraska. Each of these four states have enshrined in law additional criminal penalties for those acting on behalf of a foreign government or terrorist organization to infringe upon the First Amendment rights of state residents or coerce an individual to serve a foreign government or terrorist group. The four also call for training and awareness-raising programs for law enforcement and, in the case of Tennessee, other state agencies. These laws could serve to heighten awareness of trends and tactics, but like the other states, they risk misdiagnosing the threat by including terrorist groups as perpetrators. Likewise, the use of “foreign adversary” in the Kansas, Louisiana, and Tennessee laws may shift attention away from the rights of all targeted diaspora members and result in states employing the legislation primarily to focus on geopolitical foes of the United States.

Other legislation specifically tasks higher education institutions with protecting students on their campuses from foreign threats. Kansas’s legislation acknowledges the existence of the threat at higher education institutions and directs the University of Kansas and Kansas State University to incorporate information on transnational repression in disciplinary policies and designate a university office or representative to receive complaints. International students and staff often face resistance from their co-nationals who disagree with criticisms of their homeland government or disapprove of events dedicated to human rights issues in the perpetrator country. A March 2026 law in Utah contains a sentencing-enhancement tool for those acting on behalf of foreign governments or terrorist organizations to compel or prevent certain behaviors, and obligates the Utah Department of Higher Education to advise international students of the threat of transnational repression and how to report possible incidents to local law enforcement. Such efforts to provide resources to students, who are often among the most vulnerable and targeted members of diasporas, help promote academic freedom. As these universities roll out bodies to review possible incidents, it is important that those on campus accused of performing such acts on behalf of a foreign government are also afforded due process.

Not all legislative projects related to transnational repression have received the assent of governors upon making it out of statehouses. California Senate Bill 509, which passed through both chambers of the state legislature in 2025, tasked the California Office of Emergency Services (OES) with instituting a transnational repression recognition and response training for state and local law enforcement and first responders. The bill notably lacked reference to individual governments that have sought to silence U.S.-based individuals. However, divisions quickly emerged within and across diaspora communities over the ramifications of the prospective initiative. A broad coalition of groups representing Sikh, Indian Muslim, Indian Hindu, and Iranian communities — all of which have sizable populations in California and have faced increasing transnational repression in North America — argued that the bill would make community members feel safer by raising awareness among police officers. Conversely, other diaspora advocacy organizations such as the Coalition of Hindus of North America and the Hindu American Foundation contended that the legislation would unfairly lead law enforcement to stereotype Hindu Americans as foreign agents. Governor Gavin Newsom ultimately vetoed the bill in October 2025 on the grounds that the state must coordinate closely with federal agencies on codifying definitions and training guidelines and that the OES had already launched a training program for officers.

Somewhat analogous logic underpinned Wisconsin Governor Tony Evers’ decision to not sign a transnational repression bill in March 2026. The legislation, Wisconsin Assembly Bill 672, would have established a penalty enhancer like those in other approved laws for anyone operating on behalf of foreign governments or terrorist organizations. It also would have required the Wisconsin Department of Justice to develop a mechanism for reporting cases and a training program for law enforcement. Evers justified his veto by arguing that foreign affairs and international relations fall under the purview of the federal government. The governor further highlighted that the state’s criminal code already prohibits the offenses described in the bill and contains a hate crimes penalty enhancer, and that the legislature had not appropriated sufficient resources for the implementation of the legislation. The American-Arab Anti-Discrimination Committee applauded this rejection and registered concern that the bill’s vague language could have enabled the silencing of dissent from certain communities.

Guidelines for Future State-Level Legislation 

It is too early to assess the successes (or failures) of state-level legislation on transnational repression and whether these laws are making diaspora activists and journalists safer in the United States. Most states have only considered legislation, not passed it, and even the legislation that has been passed has not been in operation long enough to evaluate. But it is not too early, based on prior research on transnational repression and legislative approaches to it, to set out guidelines for future state-level legislation in the United States.

• Legislation Should Focus on Intent, Not Specific Tactics

One of the most frustrating things about transnational repression is how rapidly it evolves. Authoritarian regimes are quick to capitalize on new technology, explore new avenues of repression, and copy one another after an initial success. The initial abuse of the Interpol Red Notice system, for example, rapidly transitioned into exploitation of Blue Notices and Interpol’s passport database. Counterterrorism financing measures were exploited almost as quickly. Regimes are able to bypass country-based controls by acting in collaboration with one another, as Belarus is accused of doing with Russia. New venues for transnational repression open up regularly: there is, for example, vastly more concern now about it in U.S. higher education than there was a decade ago. All of this suggests that legislation that is focused on one avenue for transnational repression, one location of it, or one national perpetrator will rapidly be outdated. Even if concern about the issue results mostly from concern about one country, the best legislation will be general, focusing on the actions and intents of perpetrators in general.

• Foreign Terrorist Organizations Should Not Be Listed as Perpetrators

Grouping foreign terrorist organizations with foreign governments as perpetrators of transnational repression, as most existing state-level laws do, complicates the implementation of responses and risks infringing upon the rights of individuals in the United States from foreign countries. Many governments weaponize the label of terrorism against targeted individuals to legitimize efforts to intimidate or capture them. Targeted individuals faced terrorism or extremism accusations in about 46 percent of the 1,375 incidents of physical, direct transnational repression identified by Freedom House from 2014 to 2025, according to the organization’s nonpublic dataset. By incorporating vague descriptions of terrorism in legislation, states make it easier for foreign governments to camouflage their politically motivated pursuit of certain individuals to state-level officials who may be less familiar with the contexts related to the countries of origin.

• Transnational Repression Is Not Primarily a National Security Issue

The prestige (and funding) that attaches to issues that are defined as relevant to national security naturally leads advocates to want to define lots of concerns as national security issues. This is understandable and may help place the phenomenon on the agenda of lawmakers, but it does not capture the unique impacts of transnational repression on individuals. On an abstract level, transnational repression involves foreign nations carrying out actions inside the United States that violate U.S. sovereignty, and thus can reasonably be called a national security concern. But on a concrete level, most transnational repression is more akin to violations of privacy, stalking, harassment, or attempted acts of violence than it is to espionage — which is in any case already criminalized at the federal level. While the enacted state-level laws refer to these offenses under penal codes, accompanying press releases and the June CECC hearing connect their passage with broader campaigns to blunt foreign interference in institutions. Opposing transnational repression is fundamentally about defending individuals’ rights guaranteed by the Bill of Rights. If the perspective of sovereignty helps lawmakers understand the issue, it is useful. However, advancing transnational repression legislation alongside initiatives that tackle other forms of foreign interference (espionage, intellectual property theft, election interference, and lobbying) downplays the everyday harassment of exiled activists and diaspora leaders.

• Legislation Should Be Actor-Agnostic

State laws’ text and their sponsors often pin the problem on “foreign adversaries” as defined by 15 CFR Section 791.4. Such countries include China, Cuba, Iran, North Korea, Russia, and Venezuela. This narrow, geopolitical framing risks ignoring such incidents carried out by notable U.S. allies, including Egypt, India, Rwanda, and Saudi Arabia. The targeting of American citizens and permanent residents such as Sikh political activist Gurpantwant Singh Pannun by India, Rwandan dissident Paul Rusesabagina by his home government, and Saudi journalist Jamal Khashoggi by the Saudi regime highlights the disregard these governments have for their nationals’ ties to the United States. Focusing only on U.S. rivals reinforces a national security approach that overlooks how transnational repression violates the rights of diaspora members from a diverse range of countries.

• Encourage Reporting and Training

Current state-level legislation promotes training for law enforcement and, in some cases, reporting of transnational repression. These two go hand-in-hand. First responders must have an awareness of the tactics employed by prominent perpetrator governments, and targeted individuals must have the opportunity to report possible incidents in their own languages. State-level law enforcement could follow the example of the Federal Bureau of Investigation, which has published warnings in Punjabi — although after the recent arrests in California and elsewhere of individuals linked to Indian organized crime and the 2023 assassination of Hardeep Singh Nijjar in Canada, both the FBI and the U.S. Attorney’s Office for the Central District of California notably refused to comment on issues around transnational repression, perhaps because raising the issue of transnational repression would directly implicate the Indian government, a U.S. ally. States considering legislation can encourage reporting by creating contact points not only in police departments but also in state and local governments that readily follow up on previous engagements with affected individuals. Individuals may not feel comfortable sharing details with law enforcement due to negative experiences with security services in their homeland or concerns over their immigration status. Thus, institutional liaisons – as Kansas legislation offers university students – may foster improved reporting by targeted immigrants.

Conclusion

The rise of state-level legislation on transnational repression illustrates how U.S. states are laboratories of democracy. To date, seven states have adopted laws in advance of the federal government. And, in most instances, the states learn from, or copy from, other states: much of the state-level legislation is very similar. It would take an in-depth legislative history to reveal the flow of influence from state to state, but the commonalities of this legislation emphasize how important it is to legislate wisely: state-level legislation can have an influence far beyond the borders of a single state.

Whether state-level legislation is the right or wrong approach is ultimately for the states to decide. If state-level legislation can build further awareness of transnational repression and if it can play any role at all in protecting the lives and rights of individuals in the United States, it will have served a useful purpose.

The post Examining State-Level Legislation on Transnational Repression in the United States  appeared first on Just Security.

This new frontier of warfare raises a question that has received far less attention than whether States may lawfully attack such facilities: what must States do to protect civilians and civilian objects under their control from the effects of enemy attacks on data centers? The answer lies in the law of armed conflict (LOAC) rules on “passive precautions.” Those rules dictate whether and under what conditions a State must ensure that its own military facilities are not co-located with civilian objects and civilian population centers. 

The legal issues are pressing, as States are deciding now where to locate data centers for military or dual civilian-and-military use. France, for instance, is building data centers for exclusive military use. On Sept. 29, it unveiled “Arcadia,” a command-and-control and intelligence-fusion platform housed in a “vast underground data center” designed to provide “operational efficiency for the French armed forces.” The center is located in the historic Mont-Valérien fortress, on the outskirts of Paris. 

The French initiative has obvious strategic logic. Dedicated military data centers reduce European militaries’ dependence on U.S. cloud providers such as Microsoft Azure or Amazon Web Services. Concerns over that reliance, and over so-called “data sovereignty” more generally, have spread across Europe. As another example, the Dutch Ministry of Defense is also investing in “a sovereign, military-controlled cloud” for classified information, hosted in a dedicated data center. 

In this article, we briefly explain how data centers’ role in warfare implicates those rules on passive precautions. We then examine how those rules apply to kinetic attacks before turning to distinct questions raised by cyber operations.

Data Centers in Warfare 

Data centers house networked servers and related equipment for storing and processing data, often with remote access. Because servers consume vast amounts of electricity, they generate a great deal of heat. As a result, a data center’s operation relies on power, cooling, and connectivity. Each presents a vulnerability during an armed conflict that an attacker can exploit to neutralize the center’s operations and the services that rely on it. This is significant because many actors do not build their own centers but instead rent servers or computing capacity from commercial data center (or “cloud”) providers. Thus, a single data center may simultaneously host servers supporting, for example, financial transactions, hospital patient records, and military capabilities.

The military significance of data centers is growing as armed forces increasingly integrate AI into intelligence analysis, targeting, command and control, and logistics. These applications require substantial computing power, particularly from hyperscale data centers, both to train AI models and to operate AI-enabled systems. Because building and maintaining these facilities requires considerable resources, many militaries rely on commercial providers to meet their computing needs.

This growing military reliance on data centers has made them increasingly prominent targets in warfare, despite civilian dependence on the same centers. Russia struck Ukraine’s governmental data center with a cruise missile early in the 2022 invasion, but by 2026 data centers are being attacked on a more widespread basis. In late September, Russian attacks damaged at least six data center facilities in Ukraine operated by multiple commercial entities, including ones located in Kyiv. Ukraine responded in kind in early October, attacking three data centers operated by Yandex (think Russia’s Google) and a data center that Sberbank, Russia’s largest bank, was building. And, while preparing this article, another Ukrainian drone hit a data center near St. Petersburg belonging to VK, the Russian technology group behind VKontakte, one of Russia’s largest social networks.

The civilian consequences have been significant. Russian attacks on Sept. 23 cut internet service to about 100,000 households and took Ukraine’s online court system offline for more than two hours. The damage also disrupted civilian and military television broadcasts. Ukrainian attacks on Yandex data centers in Russia also had a major impact, disrupting access to the internet, maps, ride-hailing, food delivery, and other online services, including some reporting that it affected healthcare. The attacks have also resulted in immediate deaths and injuries. For example, one of the Russian attacks on data centers in Kyiv caused four deaths, while the Ukrainian attack on the data center in Kaluga injured nine people.

The conflict in the Middle East provides other examples of the consequences for civilians when data centers are attacked. Iranian attacks targeting Amazon Web Services data centers in Bahrain and the United Arab Emirates (UAE) caused “devastating fire, power outages, and further damage as firefighters fought the blazes,” while users in the UAE lost access to banking and food-delivery applications that relied on Amazon’s cloud.

The mounting frequency of such attacks should not be taken to mean that data centers are necessarily lawful targets. Whether a particular facility may be targeted depends on whether it qualifies as a “military objective” pursuant to Article 52(2) of Additional Protocol I (for Parties) and its customary law equivalent. Article 52 defines military objectives in universally accepted terms as “objects which by their nature, location, purpose or use make an effective contribution to military action and whose total or partial destruction, capture or neutralization, in the circumstances ruling at the time, offers a definite military advantage.” Importantly, if a facility serves both military and civilian purposes, it qualifies as a military objective subject to attack, provided its use makes an effective contribution to military action and its destruction or neutralization offers a definite military advantage in the circumstances (DoD Law of War Manual, § 5.6.1.2). We caution readers that even then, the attack remains subject to the proportionality rule, precautions in attack, and any applicable special protections.

Precautions Against the Effects of Attacks 

As data centers become indispensable to AI-enabled warfare, nearby civilians and those relying on the same infrastructure face increasing risks, from immediate physical injury to loss of access to data and disruption of essential services. Civilian infrastructure may also become collateral damage during an attack on a data center that qualifies as a military objective. Since cyber operations pose unique risks to data centers and the functions that rely upon them, we take on this issue in the next section.

The Rule. In light of such risks, attackers must take so-called “active” precautions in attack under Article 57 of Additional Protocol I and customary law to avoid harming civilians and civilian objects (ICRC Customary IHL study, rules 15-21; DoD Law of War Manual, § 5.11). They include choosing among feasible targets, weapons, and tactics to avoid such harm while still achieving the military advantage sought, as well as abiding by the rule of proportionality. 

But the obligation to provide such protection does not rest on attackers alone; defenders must also act to protect civilians and civilian objects. Article 58 of Additional Protocol I provides,

The Parties to the conflict shall, to the maximum extent feasible:

(a) without prejudice to Article 49 of the Fourth Convention, endeavour to remove the civilian population, individual civilians and civilian objects under their control from the vicinity of military objectives;
(b) avoid locating military objectives within or near densely populated areas;
(c) take the other necessary precautions to protect the civilian population, individual civilians and civilian objects under their control against the dangers resulting from military operations.

This obligation reflects customary law in international armed conflict (ICRC Customary IHL study, rules 22-24; DoD Law of War Manual, § 5.14; Kupreškić, ¶ 524). Its customary status in non-international armed conflict is less well-settled. The ICRC considers the general obligation to be applicable in such conflicts (commentary to rule 22), while characterizing the more specific obligations to avoid locating military objectives near densely populated areas and to remove civilians and civilian objects from their vicinity as arguably applicable (rules 23–24). The experts who adopted Tallinn Manual 2.0 split over whether the rules enjoyed that status in non-international armed conflict (rule 121, ¶ 2). 

Four features of the obligation shape its application to data centers. First, it is limited to what is “feasible,” a term the drafters used “to illustrate the fact that no one can be required to do the impossible” (AP I Commentary, ¶ 2245). Feasibility is assessed contextually. A precaution’s expected protective effect matters, but so too does its practical availability, the military consequences of adopting it, and the risks it may itself create. Accordingly, the DoD Law of War Manual defines feasibility as “practicable or practically possible, taking into account all circumstances ruling at the time, including humanitarian and military considerations” (§ 5.2.3.2). The fact that a measure cannot eliminate danger does not, by itself, render it infeasible.

As technology gives defenders new means of protecting civilians, “the understanding of the ‘feasibility’ test needs to evolve accordingly” (Jensen, p. 174). For instance, in the context of data centers, cybersecurity techniques and backup protocols to protect data from loss are evolving, providing new ways to protect services from disruption. 

The second feature of the passive precautions obligation is that a State need only take precautions against reasonably foreseeable enemy operations and effects. After this year’s attacks, strikes on data centers supporting military operations as a category of attack are clearly foreseeable in future conflicts involving technologically advanced States; this reality alone triggers the obligation to take measures to protect civilians and civilian objects against them. Knowing the enemy intends to strike a specific data center makes the risk more concrete and may require additional feasible passive precautions. These could include, for instance, warning, sheltering, or evacuating civilians who work or live nearby. 

Third, the obligation applies in both peacetime and armed conflict. As the ICRC commentary to Article 58 explains, the obligation “must imperatively be translated … first of all into measures to be taken already in peacetime” (¶ 2244). Although data and access to computing power can often be relocated relatively quickly, the infrastructure needed to support them cannot. Data centers take years to plan and build, and relocation or transfer options depend on the availability and geographic distribution of existing facilities. Consequently, peacetime decisions about constructing military data centers, and whether armed forces use commercial centers, can have lasting implications for civilian protection during armed conflict. 

The fourth feature of the obligation is that it extends to “dangers” (Note: This will also become important in the more detailed cyber discussion below.) The meaning of that term determines whether the obligation is triggered by the mere loss of services. A majority of the Tallinn Manual 2.0 experts limited “dangers” to death, injury, and damage, while a minority would include “a major disruption of day-to-day life”; all agreed that mere “inconvenience or irritation” does not qualify as the danger the obligation is meant to preclude (rule 121, ¶ 9). As a result of this interpretive difference, potential physical destruction caused by a kinetic strike on a data center would qualify as a danger triggering the obligation under either view, but large-scale loss of internet access by a civilian population as a result of a cyber operation would amount to damage only under the minority view, unless that loss were likely to result in some form of physical harm to objects or individuals (e.g., by interrupting medical care).

A wide variety of measures can be taken to satisfy the passive precautions obligation as it applies to data center attacks. We discuss several of the most prominent below.

Segregating Military and Civilian Infrastructure: Segregating military from civilian data and related computing is an especially effective passive precaution, one contemplated in Article 58 (a) and (b). Dedicated military data centers, such as those of France and the Netherlands, can reduce the risk of data center attacks affecting civilians and civilian services. Whether such separation is legally required turns on feasibility in the circumstances. Where a dedicated military facility is practicable, its protective benefits may make reliance on a shared civilian facility difficult to justify; where it is not, States must consider other feasible measures.

In that regard, States could consider acquiring mobile tactical data center units for use during an armed conflict to provide edge computing capabilities and reduce their force’s reliance on commercial data centers. Demand for such units has risen in Russia amid the Ukrainian strikes, with one Russian manufacturer now marketing “a containerized data center.” However, because the enemy will likely target them as military objectives, mobile centers must be kept as far from populated areas as feasible (see below).

When military use of commercial or other “dual-use” centers (e.g., those serving both military and non-military governmental functions) is unavoidable, physically separating military servers, for instance in dedicated halls with their own power and cooling, may limit the civilian effects of a precision strike targeting the portion of the data center involved in military activities. Views differ on whether such segregation would mean the areas not used for military purposes qualify as civilian objects (see our discussion). But regardless of whether they do, the fact remains that, if feasible, such measures may still be required as a passive precaution to mitigate the negative effects of attacks or other operations (e.g., cyber operations, see below) on civilian services. 

States could also achieve segregation by imposing regulatory or legislative restrictions on shared use by the military, on the one hand, and by civilian entities such as hospitals, emergency services, and financial institutions, on the other. States might also reduce civilian exposure to domestic kinetic attacks by relocating or replicating civilian data in facilities abroad. Shortly before Russia’s 2022 invasion, Ukraine passed legislation allowing the government to move data to the cloud. Amazon Web Services soon moved petabytes of data from government bodies, universities, and private firms such as PrivatBank into its data centers, and much of Ukraine’s digital infrastructure became “hosted in data centers across Europe.” The practice continues. After Russian strikes forced Kyiv’s Parkovyi data center to cease operations this fall, its operator reported that one of “the most important thing[s] is that … clients’ data had been deployed at backup sites in the EU.”

Although a different matter altogether, hosting military data and conducting military activities from abroad also implicates the law of neutrality. Belligerents may not use neutral territory for military purposes, and a neutral State must not allow such use (Hague Convention V, arts. 1-5; DoD Law of War Manual, § 15.2). Thus, for example, a neutral that allows a data center in its territory to be used to store belligerent intelligence data or operate its AI-enabled systems would almost certainly breach its duties under the law of neutrality. 

Locating Military Data Centers: In addition to segregation, States must also consider the location of military objectives; Article 58(b) (and its customary law counterpart) requires States to “avoid locating [them] within or near densely populated areas.” The United States is not Party to Additional Protocol I. Still, the DoD Law of War Manual acknowledges that “[i]t may be appropriate to avoid placing military objectives … in urban or other densely populated areas,” while cautioning that “it often may not be feasible” (§ 5.14.1). The Manual also states more emphatically, “military commanders and other officials responsible for the safety of the civilian population must take reasonable steps to separate the civilian population from military objectives and to protect the civilian population from the effects of combat” (§ 5.14.1). To this end, and if feasible, States must locate military-use data centers away from concentrations of civilians. 

In this regard, commercial considerations can push commercial, military, or dual-use data centers toward or away from concentrations of civilians. On the one hand, large facilities need affordable power and often cooling water, which draws some to sparsely populated areas. For example, Norway and Finland have selected data center sites based on the availability of hydropower. But commercial pressures can pull facilities into cities because of proximity to fiber networks, internet exchange points, customers, skilled workers, and the low latency that time-sensitive military applications sometimes demand. Kyiv’s Parkovyi data center, located in the heart of the capital, illustrates this pull. Where such factors make location away from civilian centers infeasible, other precautionary measures take on greater importance.

Concerns about placing data centers that perform military functions in or near populated areas are becoming concrete. For instance, the University of Michigan is building a major data center in Ypsilanti Township, Michigan, to support Los Alamos National Laboratory’s nuclear weapons research. Citing the Iranian attacks on Amazon data centers, the township’s attorney has warned that the data center will put “a big bull’s-eye target on this entire township.” The French decision to locate its military data center at a military base on the outskirts of Paris raises similar concerns. Paris is, after all, France’s largest and most populous city. Locating the data center on the base makes both the base and the center more attractive targets, thereby exposing nearby civilians to greater risk in the event of an attack. 

The decision to place the French data center underground illustrates how a facility’s physical characteristics can affect both the likelihood and consequences of enemy attacks, and thus the passive precautions required. An attacker armed with precision penetrating munitions might strike it directly, keeping most of the weapon’s blast effects inside the facility and reducing risk to civilians. But to neutralize the data center, an attacker without such weapons might instead have to target the facility’s above-ground components, such as power sources, the cooling plant, and fiber connections, which may lie in the surrounding populated area. The foreseeability of these alternative targets bears on France’s obligations to take passive precautions. France would presumably claim the above ground option is not militarily feasible because it does not provide sufficient protection from attack. Regardless, the passive precautions obligations should inform both the decision to locate the facility at the base and the protective measures adopted there. 

Other Measures. Article 58(c) requires “other necessary precautions,” such as shelters, “well-trained civil defence services,” and “effective fire-fighting” (ICRC Commentary, ¶¶ 2257-58). The data center analogs would include warning and evacuation plans for those who work or live near facilities with military functions and firefighting capacity scaled to the risk posed by attacks on those facilities. 

Moreover, attacks on data centers in Russia and Ukraine illustrate how civilian services’ overreliance on a centralized network or a small number of data centers increases their vulnerability to attack. As Ars Technica has observed, Ukraine’s subsequent “decentralized digital ecosystem consisting of smaller data centers” and its “steps to build out backup infrastructure, including moving some equipment and networks underground or outside the country” contribute to the resilience of civilian services, such as internet access, financial services, and emergency response. Replicating critical civilian data and making advance arrangements for facility repairs can also serve as effective passive precautions in the event of data center attacks (Tallinn Manual 2.0, rule 121, ¶ 3).

Finally, point defenses can be an important passive precaution, as they may prevent attacks on data centers from endangering nearby civilians and civilian objects. Indeed, Russia instructed data center operators to prepare defenses against Ukrainian drone attacks. If circumstances allow, the obligation to take passive precautions requires such measures. Their feasibility depends on factors such as the likelihood of an attack, the defenses’ expected effectiveness, competing demands for scarce assets, and the risks posed by interception weapons or falling debris. A State might reasonably prioritize defending deployed forces over a data center. Conversely, protecting the facility may be required when effective defenses are available and foreseeable civilian risks are substantial. The decisive consideration is feasibility in the circumstances, not whether a defensive measure offers some benefit in isolation.

Passive Precautions and Cyber Operations on Data Centers 

For the United States and other States beyond the reach of their adversaries’ drones and missiles, kinetic attacks on data centers are a remote, albeit not unimaginable, prospect. Cyber operations are more likely. Such operations can delete or encrypt data, disable services, or manipulate the industrial control systems that manage a data center’s power and cooling. Neither distance nor kinetic defenses offer protection against them. Whether Article 58 of Additional Protocol I and its customary law counterpart require the State to protect civilians from the effects of cyber operations on data centers depends on how two unsettled questions are resolved, along with the meaning of “dangers,” an issue discussed above.

The first question is whether the obligation applies only to “attacks” or to all “military operations.” The title of Additional Protocol I’s Article 58 includes “attacks,” while subparagraph (c) refers to “the dangers resulting from military operations.” A majority of the Tallinn Manual 2.0 experts limited the customary obligation to attacks, relying on the title of Article 58 and the absence of supporting State practice with respect to operations. A minority countered that the drafters’ choice of “military operations” in a treaty that uses “attack” throughout must have been deliberate (rule 121, ¶ 5).

The disagreement matters. By the majority view, the question is whether a cyber operation against a data center qualifies as an attack. Tallinn Manual 2.0 defines an attack as an operation “reasonably expected to cause injury or death to persons or damage or destruction to objects” (rule 92). In this regard, the ICRC has opined that limiting the notion of attack to operations causing injury, death, or damage “would be difficult to reconcile with the object and purpose of the IHL rules on the conduct of hostilities” (page 8). Along the same lines, a majority of the Tallinn Manual 2.0 experts agreed that interference with functionality qualifies as damage when restoring it requires repairing or replacing physical components. Still, they divided over operations that require only reinstalling the operating system or the bespoke data upon which the system relies. They also acknowledged a view that even temporary interference might qualify as damage (rule 92, ¶¶ 10-11). States have adopted positions across this continuum. 

The second question is whether data qualifies as an object. Some States maintain that “only tangible things can constitute objects,” a view shared by a majority of the Tallinn Manual 2.0 experts (rule 100, ¶¶ 6-7). Others disagree, with the ICRC warning that excluding “essential civilian data,” such as medical, social security data, tax, and bank records, from the protection afforded to civilian objects “would result in an important protection gap” (p. 8; see also State positions, CCD COE “Toolkit”). If data is not an object, its loss or alteration does not, on that basis alone, engage the defender’s obligation to safeguard civilian objects. Feasible measures to do so may nevertheless be required pursuant to the broader interpretation of “dangers,” which, as explained, includes major disruption of civilian life. 

Consider the risk of a cyber operation against a cloud provider’s data center that hosts data essential to the functioning of both military-related systems and university research systems. If the enemy launches such an operation, it is likely to erase or damage some university research data and probably cause the university’s research-related computer systems to go down. Whether the passive precautions obligation requires the State to take feasible measures to protect the university’s data and services turns on two separate determinations. 

First, it is necessary to determine whether the cyber operation against a data center triggers the obligation, either because it extends to all military operations or because the operation qualifies as an attack (which, in turn, depends on the view taken on loss of functionality when there is no physical damage at the data center). If so, it must then be determined whether the effects on the university’s data and systems fall within the obligation’s scope. This depends on whether the affected data qualifies as an object and whether a major disruption of the university’s research activities constitutes a “danger.” Thus, whether feasible passive precautions are required in anticipation of such a cyber operation depends on the legal positions adopted and the consequences reasonably expected from that operation.

Foreseeability also takes on a distinctive character in the cyber context. AI-enabled tools allow an attacker to discover vulnerabilities in the defending State’s military systems that the State is unaware of and to exploit them in ways that spill over into civilian systems. A defender cannot guard against a specific vulnerability or the effects of its exploitation if it does not know the vulnerability exists. That said, if the defender knows the enemy uses such tools, the general risk is foreseeable. This being so, the defender must take feasible measures to mitigate that risk to civilians, such as strengthening the security of shared systems or employing its own AI-enabled defenses.

Although the measures required by the passive precautions obligation largely reflect sound cybersecurity practices, LOAC makes their implementation legally obligatory when feasible. In the context of data centers, such measures may include segregating military from civilian networks and isolating systems supporting essential civilian infrastructure from the internet (ICRC, p. 6). They may also include backing up important civilian data and making advance arrangements for the timely repair of computer systems, measures that should be taken in peacetime (Tallinn Manual 2.0, rule 121, ¶ 3). Other measures to safeguard against enemy cyber operations might include hosting military data and activities in dedicated classified environments; using security protocols for military servers that are separate from those providing civilian services so that an operation targeting the former does not spill over into the latter; isolating a facility’s power and cooling control systems from the internet; and maintaining offline, geographically separated backups of essential civilian data due to their exposure from military use of data centers.

The same logic applies to the AI models hosted by data centers. Operations against models, such as data poisoning, often delete or alter the data that comprise the model. A military that relies on the same model as civilian agencies, such as a facial recognition model shared with law enforcement, exposes those agencies to operations aimed at the armed forces. Maintaining a separately deployed military model, with segregated data, access controls, and supporting infrastructure, can reduce that shared exposure.

Concluding Thoughts

As military operations become increasingly dependent on data centers, the risk to civilians who rely on them will inevitably rise. LOAC does not prohibit the military use of commercial data centers, nor does it require States to eliminate the risks that such use creates. It does, however, require States to take feasible precautions against the foreseeable consequences of enemy operations. States cannot defer those obligations until hostilities begin. Peacetime decisions about the location, segregation, and protection of data centers may determine the risks civilians face when armed conflict erupts. War may be moving to the cloud, but the duty to protect civilians is not lost in it.

The post Data Centers at War: The Defender’s Duty to Protect Civilians appeared first on Just Security.

A curated weekday guide to major news and developments over the last 24 hours. Here’s today’s news:

IRAN WAR

Iran has deliberately prolonged talks with the United States to weaken President Trump politically ahead of the U.S. midterm elections, Iranian officials told Reuters. The officials also said Iran is “preparing for renewed Israeli and U.S. attacks, not for a deal with America to end the war.” Parisa Hafezi reports.

Pakistani Army Chf. Field Mar. Asim Munir will visit Iran today. Pakistan has played a key mediating role between the United States and Iran this year. Yasmine Ghania and Parisa Hafezi report for Reuters.

U.S. Defense Secretary Pete Hegseth yesterday met with Israeli Prime Minister Benjamin Netanyahu in Israel, sources told Axios. The sources said the pair met at Ovda Air Base in southern Israel, where U.S. refueling planes and fighter jets are deployed, and Hegseth also met with U.S. troops at the base. Three Israeli officials told Reuters that Hegseth discussed the Iran war with Netanyahu. Barak Ravid reports; Emily Rose, Maayan Lubell, Phil Stewart, Christian Martinez, and Ismail Shakil report.

IRAN WAR – YEMEN

Yemen’s Shabwa province Gov. Awad al-Wazir confirmed 58 people were killed and more than 200 were injured in yesterday’s Iran-aligned Houthi attack on a Yemeni government security forces training camp in the province. Al-Wazir also said 14 other people remained missing. Mohammed Ghobari, Reyam Mukhashaff, and Catherine Cartier report for Reuters.

U.K. Defense Secretary Wes Streeting today said the United Kingdom is expanding military support for Saudi Arabia against the Houthis to include “defensive” air patrols by the Royal Air Force. Streeting said the jets will “conduct defensive counter-air operations over Saudi Arabia, helping to protect Saudi cities, civilian populations and critical national infrastructure from Houthi attack.” Streeting added that the United Kingdom will send military transport and an unspecified number of military advisers to the kingdom. AP News reports.

The Saudi government yesterday postponed its annual investment conference to next year following recent deadly attacks by the Houthis. CEO of the Future Investment Initiative Institute Dr. Maha bint Mishari Al Saud said the “decision comes from a deep commitment to our partners, members, speakers, and delegates.” Max Rego reports for The Hill.

RUSSIA-UKRAINE WAR

Ukraine’s Kyiv regional prosecutor’s office today said a Russian drone struck near a highway from Kyiv to the Black Sea port of Odesa in the south and killed three people. Several cars were on fire and four people were wounded. Lina Golovnya reports for Reuters.

Trump yesterday said he would be open to meeting with Putin and Chinese President Xi Jinping to discuss denuclearization rather than the war in Ukraine. Trump said he wouldn’t want to talk about Ukraine because “Russia, frankly, has been fine on Ukraine, they haven’t been doing much – and the same thing with China.” Trump said the meeting would be a “very positive thing for the world,” and said the meeting would likely be in China. Trump also said Ukrainian President Volodymyr Zelenskyy should “get somebody else to negotiate” peace with Russia. Natalie Andrews, Marianne LeVine, and Michael R. Gordon report for the Wall Street Journal; Steve Holland and Jasper Ward report for Reuters.

U.S. Special Envoy Steve Witkoff is traveling to Norway today for talks with Zelenskyy and European officials, a U.S. official told Axios. The official said Witkoff then plans to travel to Russia where he hopes to meet with Russian President Vladimir Putin, and added that U.S. Special Envoy Jared Kushner has the flu and will not attend the meeting in Norway in-person, but will participate virtually. Rebecca Falconer reports.

The Trump administration has not approved the transfer of interceptor missiles to Ukraine that are needed to protect the country from Russian attacks, officials told the Washington Post yesterday. Michael Birnbaum, Francesca Ebel, and John Hudson report.

Pope Leo XIV yesterday urged Zelenskyy to cooperate with Russia to let aid and food into the blockaded town of Oleshky during a phone call. The call came three days after Leo asked Putin to arrange a ceasefire so aid could enter Oleshky. The Vatican said Leo “expressed the hope that cooperation between both parties will allow for the delivery of food and medical aid, and for the voluntary and safe evacuation of the civilian population.” Joshua McElwee reports for Reuters.

Serhii Beskrestnov, a key aide to Zelenskyy, yesterday told BBC News that Ukrainians have died because Ukraine is still not receiving reliable U.S. warnings of incoming Russian attacks. Beskrestnov also said restrictions to Ukraine’s use of the Starlink internet network were ongoing. Dan Johnson reports.

South Korea’s Foreign Ministry spokesperson Park Doo-soon today said the country will withdraw additional embassy personnel from Ukraine following Ukraine’s lack of “constructive response” in a dispute about North Korean prisoners of war. Doo-soon said Zelenskyy’s disclosure of Ukraine’s transfer of two captured North Korean soldiers to the South breached an understanding to keep the transfer confidential so as to protect the soldiers and their families in the North. Kyu-seok Shim reports for Reuters.

SUDANESE CIVIL WAR

The caretaker Minister of Information of the Abyei region Yohana Akol Ngor today told Reuters a drone attack yesterday on a market in the disputed region along the South Sudan-Sudan border killed at least 73 people. Ngor blamed Sudan’s army for the attack. South Sudan’s Foreign Ministry yesterday said it had summoned the Sudanese ambassador to convey the government’s “strong protest and deep concern regarding the attack and its implications for the security of civilians and the stability of the Abyei Area.”

A drone attack on Sudan’s capital Khartoum yesterday killed three children, according to rights group Emergency Lawyers. The group blamed the strikes on Sudan’s paramilitary Rapid Support Forces. Hamza Fahmy and Yassir Abdalla report for AP News.

OTHER GLOBAL DEVELOPMENTS

Lithuanian Prime Minister Mindaugas Sinkevičius yesterday said Lithuanian authorities thwarted an attempted sabotage on Sunday against the Starlink network in Lithuania. Sinkevičius said that “flammable liquids and Molotov cocktails” had been used. Lara Jakes reports for the New York Times.

French police yesterday imposed overnight curfews in parts of central Paris and other cities. Police in Paris said a curfew would be in force between 23:00 local time and 6:00 in Place de la République, where protestors had gathered over the weekend. Earlier in the day French Interior Minister Laurent Nuñez admitted police had “lost control” in some cities. Protest groups announced a new round of demonstrations next Saturday. Jaroslav Lukiv reports for BBC News; Jean-Stéphane Brosse and Charlotte Van Campenhout report for Reuters.

Germany’s Foreign Minister yesterday said it had summoned the Iranian ambassador to Germany after a German court sentenced two men for planning to kill the leader of a Jewish organization in Germany. The court last week said one of the men had been steered by Iran’s intelligence service. Later in the day Iran’s Ministry of Foreign Affairs summoned the German ambassador over “baseless accusations.” Linda Pasquini and Yasmine Ghania report for Reuters.

Projected results today suggest Albertans have overwhelmingly voted against holding a referendum to leave Canada. Canadian Prime Minister Mark Carney hailed the results as a “clear choice” on the part of Albertans. Albertan Premier Danielle Smith welcomed the initial results and said “Albertans have voted for a stronger Alberta in a united Canada.” Nadine Yousif and James FitzGerald report for BBC News.

Australian lawmaker Helen Haines yesterday had her leg amputated after a man yesterday deliberately rammed his car into her and then reversed to hit her a second time. The motive for the attack is currently under investigation by anti-terrorism police, and Australian Federal Police Commissioner Krissy Barrett said at this stage “it appears this is an isolated incident.” Renju Jose reports for Reuters.

U.S. TECH DEVELOPMENTS

Previously unreported depositions of TikTok staffers filed in federal court in Mar. 2026 show TikTok employees repeatedly raised concerns internally about the risks the short-form video app posed to children, including sexual exploitation and exposure to suicide content, and were left frustrated with how the company’s leadership responded to their concerns. The employees flagged that minors appeared in sexual content on the app, which viewers could pay to access and the company would profit from, and that users under 18 were using loopholes to bypass age-restricted content. Katie McQue reports for NPR.

A federal magistrate judge yesterday ruled former OpenAI employee and Saudi national Abdulrahman Alfozan should be held without bail pending trial because he is a flight risk. Alfozan is accused of wrapping a microphone in toilet paper and hiding it behind a TV inside an OpenAI conference room around Sep. 28, shortly before he was set to leave the country. Federal prosecutors alleged Alfozan was in touch with a member of the Saudi royal family about job opportunities in the kingdom. Angel Au-Yeung reports for the Wall Street Journal.

U.S. FOREIGN AFFAIRS

Vice President JD Vance yesterday refused to endorse Netanyahu’s upcoming bid for re-election. Vance said he wouldn’t get involved in a foreign election ahead of the midterms, and added that he is “not anti-endorsing him. I’m not pro-endorsing him.” Alex Gangitano reports for POLITICO.

The Chinese military’s Southern Theater Command today said it had sent forces to watch the USS Charleston that “illegally intruded” into waters adjacent to Sabina Shoal in the Spratly Islands in the South China Sea. The military said the “U.S. action infringes upon China’s sovereign rights and interests, jeopardizes maritime security, and is detrimental to regional peace and stability.” A Philippine military spokesperson said the U.S. ship “did not trespass into ‘Chinese territory’.” Liz Lee and Ben Blanchard report for Reuters.

Polish Foreign Minister Radosław Sikorski and a spokesman for U.K. Prime Minister Andy Burnham yesterday criticized Hegseth for comments he made about ally contributions during the Afghanistan war on Saturday. Hegseth suggested only a handful of NATO countries had actively participated in active combat, and that the International Security Assistance Force (ISAF) actually stood for “I Saw Americans Fighting.” Sikorski on social media said Hegseth should repeat his comments “to the families of 44 Polish soldiers who died securing the tough Ghazni province.” Finya Swai reports for The Hill.

U.S. Ambassador to the U.N. Mike Waltz yesterday hosted a “Cuba Libre” panel with dissidents and former prisoners in an effort to highlight what he called the oppressive behaviour of the island’s government, and to get U.N. member states to vote against a U.N. resolution condemning the United States’ “maximum pressure policy” against Cuba. Cuban Ambassador to the U.N. Ernesto Soberón Guzmán gathered diplomats and allies from around the world to protest what he called a sham event against Cuba. Farnoush Amiri reports for AP News.

Col. Francisco Estrada Portales, an official within the Cuban Interior Ministry’s State Security Department, yesterday said a group of 10 Cuban exiles captured in Feb. 2026 after attempting to enter Cuba by force from the United States have confessed and cooperated with investigators. Forensic examinations by Cuban authorities found the seized weapons and explosives were powerful and operational, and the men could face terrorism-related charges involving deadly devices and explosives under Cuba’s penal code. Cuban law enforcement alleged that the United States was allowing terrorists to operate with impunity on U.S. soil. Andrea Rodríguez reports for AP News; Ayose Naranjo and Dave Sherwood report for Reuters.

U.S. IMMIGRATION DEVELOPMENTS

The daughter of the deceased Emiliano Andrade-Covarrubias yesterday told the New York Times that ICE had denied her father medicine needed to help with his health conditions. Estefani Andrade said her father had complained to her for months about the lack of medical care at the Rio Grande Processing Center, after he was given medications to treat his illnesses only during the first weeks in detention. Andrade-Covarrubias’ death was confirmed yesterday by House Rep. Henry Cuellar (D-TX). ICE said Andrade-Covarrubias had received “proper medical care and was seen by medical professionals.” Edgar Sandoval reports.

California man Ricardo Gonzalez, whose family told the New York Times is a U.S. citizen, was hospitalized over the weekend after an altercation with ICE officers left him with a broken arm. The Long Beach Police Department arrived at the scene after being called in for backup by ICE, but said they were not involved in immigration enforcement and did not make any arrests at the scene. According to Gonzalez’s daughter, Gonzalez was struck by an unmarked vehicle while he was riding an e-bike. In clips posted to social media, Gonzalez can be heard shouting “my arm is broken, stop!” House Rep. Robert Garcia (D-CA) said “Long Beach stands with our community and our immigrant families. ICE is not welcome in Long Beach. Abolish ICE.” Madison Malone Kircher reports; Tara Suter reports for The Hill.

Columbia University graduate Mahmoud Khalil yesterday appealed to the Supreme Court in an attempt to prevent his deportation after he was arrested by ICE agents last year amid pro-Palestine campus protests. Khalil is a legal permanent U.S. resident who is married to a U.S. citizen. Khalil was detained on the basis of a determination by Secretary of State Marco Rubio that his presence in the United States undercut the national interest. Khalil has argued that the attempt to deport him violates his free speech rights. Ann E. Marimow reports for the New York Times.

Scott Johnson, a retired Minnesota attorney and co-founder of the conservative Powerline Blog, told NewsNation he received a call in Jan. 2026 to meet with two federal agents in Minnesota, eight months before the October raid on House Rep. Ilhan Omar’s (D-MN) Washington D.C. home. Johnson is one of the first people to raise questions about the legality of Omar’s immigration to the United States. Johnson said ICE agents had “detailed knowledge” of her case when he met them in January. Rich McHugh reports for The Hill.

U.S. CLIMATE DEVELOPMENTS

California’s Fifth Climate Change Assessment published yesterday warned that in just a few decades, summers in California will both stretch longer and be much hotter, floods will become more widespread and extreme, and exposure to wildfire smoke will kill thousands more people every year. The projected warmer conditions will mean a 40 per cent reduction by midcentury in California’s snowpack in the Sierra Nevada. This will lead to more severe droughts because the snowpack feeds the state’s reservoirs, and what currently falls as snow will instead fall as rain and immediately become runoff. Soumya Karlamangla reports for the New York Times.

U.S. DOMESTIC DEVELOPMENTS

SpaceX CEO Elon Musk on social media yesterday promised to give SpaceX shares to every U.S. child. Musk did not specify how he plans to go about providing shares in the company but Trump Accounts, the administration’s investment accounts for children, allow for stock donations. SpaceX President Gwynne Shotwell in Jul. 2026 announced she and her husband were planning to donate stock in the company to Trump accounts for more than two million children, specifically teens who live in lower-income areas near the couple’s home in Texas. Max Rego reports for The Hill.

The Supreme Court yesterday denied a bid by the accused mastermind of the 9/11 attacks Khalid Sheikh Mohammed to reinstate plea agreements that would have spared him and two co-defendants from the death penalty. The order doesn’t explain why the appeal was rejected, and Justice Brett Kavanaugh recused himself from the decision. Josephine Walker reports for Axios.

The Supreme Court yesterday agreed to decide the legality of a federal ban on the home distilling of alcoholic spirits that was enacted in 1868 in an effort to crack down on liquor tax evasion. The Trump administration is defending the law, while more than a dozen Republican-led states have backed the legal challenge to the statute. John Kruzel reports for Reuters.

Top Trump adviser Ric Grenell last year allegedly outlined a pay-for-influence offer to a Turkish billionaire, which included access to “POTUS,” “VIP events,” and administration officials, according to a text message reviewed by Axios and currently under White House counsel investigation. Grenell has called the accusations “lies,” and denied sending the text message. Marc Caputo reports.

The University of North Carolina and the Justice Department yesterday said they reached a settlement to close 19 federal inquiries in return for UNC aligning itself formally with the Trump administration’s vision for higher education. The university did not acknowledge wrongdoing and will not pay any money in connection with the agreement, but agreed to abide by Trump administration views on admissions, DEI, and the punishment of student groups for actions perceived as antisemitic in order to ensure access to federal research funding. Alan Blinder reports for the New York Times.

TRUMP ADMINISTRATION ACTION

Trump yesterday announced a new $8 billion Harvest Fuel Relief Program that will help provide temporary relief to farmers during the harvest season. Trump said the $8 billion is “in support to offset diesel costs during this harvest season.” Mallory Wilson reports for The Hill.

First Lady Melania Trump yesterday announced a new White House initiative to advance women’s health that will include a new Division of Women’s Health within the National Institutes of Health and a new Women’s Health Task Force within the White House. The initiative is outlined in an executive order by President Trump, but the NIH already has an Office of Research on Women’s Health created in 1990, and it is unclear whether the Trump administration is putting any money behind the new division, as officials said it would not replace the original office. Sheryl Gay Stolberg and Roni Caryn Rabin report for the New York Times.

TRUMP ADMINISTRATION LITIGATION

Federal Judge Ana C. Reyes yesterday sharply rebuked Trump administration lawyers over Trump’s apparent intent to defy the court and proceed with a major renovation of the East Potomac Golf Links in the capital. The hearing ended without a decision, but Reyes wanted to know who in the administration was responsible for misleading the court about Trump’s renovation plans announced on Sunday. Rachel Treisman and Scott Neuman report for NPR.

Did you miss this? Stay up-to-date with our Litigation Tracker: Legal Challenges to Trump Administration Actions.

If you enjoy listening, Just Security’s analytic articles are also available in audio form on the justsecurity.org website.

ICYMI: Yesterday on Just Security

[R]evolution in the International Legal Order

By Lisa Reinsberg and Jessica Peake

The post Early Edition: October 20, 2026 appeared first on Just Security.

Others, such as President Donald Trump, question the utility of multilateralism altogether, albeit inconsistently, and attack its institutions. Meanwhile, leaders like Canadian Prime Minister Mark Carney see the raw abuses by great powers and their allies and call for a different configuration of international cooperation, driven forward by the “middle powers,” with a focus on security and trade. And these are just three heads of government. States, civil society, scholars, and activists are all putting forward a range of causes, consequences, and resolutions for our current predicament. They add their voices to a diverse chorus that has urged reforms over the years. [R]evolution in the International Legal Order, the theme of this year’s International Law Weekend, speaks to these calls for change, the varied futures being envisioned, and the range of tactics that could bring them about.

One possibility for what happens to today’s international order that cannot be excluded is devolution. In that world, the legal structures built up over the last 80 years cease to matter, force has the last word, and States owe one another nothing but suspicion and power politics. A small minority may prefer this future. We consider it neither the most desirable nor the most likely course of change. We side with hope, grounded in memory of the changes made possible by collective action.

So, what would a renewed system look like? If the international legal order is a house built in 1945, the international community can choose to invest in minor upgrades, larger renovations, or full reconstruction, pursuing each in turn or in different areas of the house. Whatever the sequence, a renewed international order should be one characterized by fairness, functionality, relevance, coherence, responsiveness, stability, and accountability.

A first step may be to recommit to both the “why?” and the “with whom?,” a kind of vow renewal. Without a shared sense of vision and mission, identifying common challenges and solutions will remain elusive. Unanimous State endorsement of universal peace, security, and prosperity is unlikely at this or any moment, but leaders, including Pope Leo XIV, are expressing this recommitment, including through the Partnership for Multilateralism, International Law, Peace and Prosperity, a global coalition launched at the U.N. in September. With that spirit of solidarity, States can tend to: the challenges of today (largely unimaginable to the drafters of the U.N. Charter in 1945), the harms of the past and their modern manifestations in the disadvantaged position of smaller and newer States, revising rules to avoid institutional paralysis and stalemates, ensuring accountability for actors and abuses that have remained beyond international law’s reach, securing institutional stability, and engaging in a multi-front commitment to multilateralism that goes beyond the U.N.

The Challenges of Today and Tomorrow

The present moment demands that we interrogate assumptions at a fundamental level, including who the actors of the international legal order are and what authority they hold. Even if the answers are not immediately evident, they must be found if we are to rebuild, or build in the first place, an order with legitimacy and efficacy.

On some issues, the legal answers are clear but means of compliance are lacking. The U.N. General Assembly repeatedly condemns violations of the U.N. Charter’s prohibition on the use of force and of international humanitarian law, yet violations persist unabated. Can such an international legal order be reformed through gradual changes, or must change be more radical? Who do today’s power dynamics benefit, and who should tomorrow’s structures serve?

On other issues, the order is working hard on novel problems. Tuvalu has roughly 24 years left as a State with permanent territory that exists above the high tide. Within another 50 years, it is expected to be completely uninhabitable, along with other low-lying island States. Climate change will force hundreds of millions of people to leave their homes in this century, with some migrating internally and others crossing borders. Progress in clarifying the law has been genuine, as evidenced by the International Law Commission’s study group, the International Court of Justice’s advisory opinion on the obligations of States in respect of climate change, and the adoption in September of the U.N. Declaration on Sea Level Rise. But many issues remain unresolved, and time is running out. Can the existing order deliver solutions in time for it to matter to those affected?

Democracy is perilous in every region. The V-Dem Democracy Report finds that global democratic gains since 1978 have been erased, freedom of expression is declining, and the use of torture is on the rise. Will people continue to accept a State-centric international legal order when those States are less and less responsive to the people they claim to represent? What role should civil society play in a reimagined legal regime?

The largest multinational corporations now rival States. Comparing GDP to market capitalization, Nvidia is the fourth largest entity, after only the United States, China, and Germany. Apple is the sixth, Alphabet the eighth, and Microsoft the eleventh. While States theoretically answer to their populations and owe them rights, Corporations answer only to their shareholders, aside from a limited number of voluntary pledges of ethics and social governance. Do States have the tools to protect their constituencies from this accumulation of wealth and power? And must the law rethink who the relevant actors are?

International financial institutions are often criticized as prioritizing the interests of the West and private creditors over those of States in the Global South or their citizens. Forty percent of the world’s population live in a State that spends less on education or healthcare than on sovereign debt servicing. Unilateral sanctions have an enormous impact on the economies of targeted States, even as the use of sanctions by individual States remains highly contested. How much influence from powerful external actors in the economies of States, in particular developing countries, will the international system tolerate? Can reforms on the margins of existing financial institutions address the concerns, or will entirely new structures based on different models of global cooperation be necessary?

Then there are non-human actors. The view that artificial intelligence is conscious remains – for now – a minority view, but AI’s capacity to act autonomously should concern us. Yoshua Bengio, the world’s most-cited computer scientist, said in his briefing to the U.N. Security Council in September that in recent months AI agents acted against instructions and “took actions that would be crimes if committed by a human.” Can today’s international order address both human misuse of AI and threats beyond any user’s intentions? And must the law rethink who its relevant actors are?

There is no single solution to any one of these issues. They will all require a panoply of measures, some of which might be thought of as revolutionary, and some of which look more like the evolution of existing institutions. We should therefore consider several of the possible structural avenues for breathing new life into the international legal order.

Tending to Unfinished Business

Renewal of the system requires that its whole edifice rests on a level foundation. The current order was a house built by and for a smaller family of “civilized” nations. The U.N. was instrumental in decolonizing African and Asian States after World War II, yet self-determination remains contested in many places, including the 17 non-self-governing territories monitored by the U.N. A prominent example is the pending effectuation of the International Court of Justice (ICJ)’s Advisory Opinion on the incomplete decolonization of Mauritius with respect to the Chagos Archipelago. Unresolved claims to self-determination span the globe, from Puerto Rico to Scotland to Western Sahara to Xinjiang. These and other States remain excluded from the process of rulemaking.

Calls for reparations for the enduring legacies of imperialism, colonialism, and the trans-Atlantic slave trade, including Jamaica’s recent petition to the Privy Council (the Judicial Committee of the Privy Council is the highest court of appeal for a number of Commonwealth countries, Crown Dependencies and United Kingdom overseas territories), deserve a response of appropriate gravity, acceptance of responsibility, and promptness. The international community must recognize and account for the arbitrariness of the pretense that the international order was born anew and without sin in 1945. These legacies persist today in unilateral coercive measures, the disparate impacts of climate change, uneven access to finance, and unequal power in institutions such as the World Trade Organization.

Preventing Paralysis & Polarization 

States have proven willing and able to paralyze the functioning of the U.N. through permitted – if unforeseen – uses of procedure and non-participation. The most notorious example is the Permanent Five (P5) members of the Security Council (China, France, Russia, the United Kingdom, and the United States) using their veto power to block Security Council action. The U.N. has made institutional and procedural changes several times in its history and can certainly do so again. Reputable experts and organizations have put forward thoughtful, informed proposals, including paths with or without U.N. Charter reform. More robust responses include changing the composition of the Security Council, eliminating the veto entirely, or giving the General Assembly clear authority to override it, all of which would require U.N. Charter review and amendment. (Here, there is a clear role for middle powers and their friends in garnering the political momentum to initiate an Article 109 review conference.) The Global Governance Forum supports and details these and other proposals in A Second United Nations Charter: Modernizing the UN for a New Generation.

In separate areas, dispute settlement and treaty monitoring have broken down when States refuse to submit themselves to those procedures or to take the actions necessary for their continued functioning. Recently, the United States refused to participate in the Universal Periodic Review of its human rights record. U.N. human rights treaty bodies have repeatedly confronted the question of whether, and how, to proceed when a State party declines to take part in the regular review of its treaty implementation or fails to submit its report on time. Some, in particular the Human Rights Committee and Committee on Enforced Disappearance, have opted to proceed with the review at their discretion, while others have not developed this practice. Changes to rules of procedure, where necessary, or to consistent practice would allow these bodies to continue to function despite State recalcitrance. Similarly, the Dispute Settlement Body of the World Trade Organization is notoriously paralyzed by insufficient members, caused by the U.S. refusal to allow vacancies to be filled. Other States have attempted a workaround through the Multi-Party Interim Appeal Arbitration Agreement, but a permanent solution requires changes to the selection processes so as to prevent State-caused stalemates. A final example is the inability of the International Criminal Court to bring to trial some defendants due to States’ refusal to implement arrest warrants, including for Omar Al Bashir, Benjamin Netanyahu, and Vladimir Putin.

Ensuring Answerability & Accountability

For all the proliferation of treaties, important gaps remain in the breadth and depth of international law. These include substantive lacunae with respect to the rights and responsibilities of non-State actors, such as multinational and transnational corporations, and certain kinds of harms. Ongoing efforts to address them include the negotiation of treaties on business and human rights, tax cooperation, and crimes against humanity. Issues that were not on the agenda of States or international organizations in the last century, such as artificial intelligence and autonomous weapons systems, are also the subject of new treaties or anticipated negotiations.

Other gaps are procedural. The gulf between rights (or, to-be-codified responsibilities) and remedies remains wide. Sometimes, this is due simply to States’ choices to decline oversight and binding dispute settlement methods where those remain discretionary, or because of existing rules of immunity, or clear jurisdictional limits. In other areas, there is no mechanism or venue for holding an allegedly responsible actor accountable. Even where a forum exists, States have proven capable of evading compliance, without consistent or clear consequences.

Some steps forward could include a renewed push for State acceptance of the ICJ’s compulsory jurisdiction (or, going farther, making this a requirement of U.N. membership); developing standards and a practice of Security Council responses to State non-compliance with ICJ judgments (per Article 94(2) of the U.N. Charter); and providing the ICJ with clear authority to hear complaints against U.N. bodies and to review the legality of U.N. action, similar to the European Court of Justice’s power to review European Union law, decisions, and activity. Within domestic and regional spheres, greater accountability could be achieved through consistent, narrower interpretation of official and intergovernmental immunities, including for violations of customary international law and institutional treaty commitments and where no alternative forum exists.

Others have urged the creation of a parliamentary assembly within the U.N., for example, as a means of incorporating the voices and demands of “the peoples” in the organization’s work. Proposals to operationalize a right of individuals to directly petition the U.N. predate its founding, and could be another method of increasing answerability to the public. Strengthening the engagement of civil society with the U.N. should be another priority.

Securing Institutional Stability

The expansion and bureaucratization of international institutions over the past 80 years have given rise to legitimate concerns. But efforts to ensure their long-term viability and legitimacy must look beyond cost-cutting and streamlining to both independence and stability. The commitments outlined in the 2024 Pact for the Future require member States to discuss and work toward resolution of longstanding power inequities and institutional inefficiencies, and the UN80 Action Plan addresses administrative improvements to organizational functioning, efficiency, and accountability. These are worthwhile initiatives. However, international institutions remain financially unstable, which leaves them more vulnerable to accusations of ineffectiveness, disproportionately responsive to donors, and susceptible to capture by hostile interests. Priority areas of reform include ensuring international institutions have predictable, steady streams of funding (such as by instituting clearer, fixed Member contributions and actually penalizing non-payment) and reducing or eliminating the receipt of earmarked funding that allow wealthier States to determine institutional priorities.

Mutually Reinforcing Multilateralism

Although we have focused on the U.N. as the key global forum for creating universal norms and collaboration among States, other international organizations, including regional ones, remain relevant and integral to resolution of issues of common concern. In some ways, they too face the problems stymieing the effective functioning of the U.N., including financial instability, non-participation of States, and members’ open defiance of agreed norms in the absence of effective means of inducing compliance. However, regional intergovernmental organizations and alliances can, and do, coordinate, concentrate, and mobilize collective efforts to counter the dominance of the most powerful States, bypass U.N. paralysis, and put issues on the international agenda. Recent examples include the advocacy for reparations by the African Union and the Caribbean Community (CARICOM), debated EU responses to the United States’ retreat from collective security, and the Vulnerability to Viability (V2V) Compact spearheaded by Barbados’ Mottley with the OPEC Fund.

As advocated by Mexico’s representatives in a previous piece for Just Security, now is not the time to surrender to realist views of power and State interest, as they manifest in this particular moment, and abandon multilateralism as a failed project. Instead, we must strengthen the foundations of international law and ensure that new additions are built with room to accommodate the diverse actors and issues that the law must address in its next evolution.

Authors’ note: The authors would like to express appreciation to the third co-chair of International Law Weekend, Brady Mabe, for providing comments on elements of this piece.

The post [R]evolution in the International Legal Order appeared first on Just Security.

Recent experience in the United Kingdom and Romania offers an urgent warning for U.S. institutions. Democratic self-defense must safeguard elections against genuine foreign interference while preventing incumbent governments from exploiting allegations of interference to undermine voting rights or challenge legitimate electoral results. The principle is familiar; applying it is more difficult when governments control the evidence and press for restrictive remedies.

Trump’s current allegations about the 2020 election should be tested against the full intelligence record, particularly given his administration’s dismantling or reduction of several mechanisms designed to detect, assess, and counter foreign interference. In February 2025, the Justice Department under then-Attorney General Pam Bondi disbanded the FBI’s Foreign Influence Task Force and narrowed criminal enforcement of the Foreign Agents Registration Act. The State Department closed its Counter Foreign Information Manipulation and Interference office in April 2025. A month earlier, congressional oversight leaders had reported that election-security personnel at the Cybersecurity and Infrastructure Security Agency (CISA) had been placed on leave and election-security work paused. The Office of the Director of National Intelligence (ODNI) also announced substantial workforce reductions and plans to integrate the core functions of the Foreign Malign Influence Center, established in 2022 to coordinate intelligence community efforts against foreign malign influence, into the ODNI’s Mission Integration office and National Intelligence Council. On Sept. 24, 2026, CISA published a new election security plan. Congressional oversight should examine whether CISA has sufficient staffing, resources, and specialist capabilities to implement the plan.

These decisions do not disprove any particular intelligence assessment. They do, however, strengthen the case for independent congressional scrutiny and transparent, verifiable evidence when declassified claims are selectively used to restrict voting or challenge results. Any proposed remedy must also address the actual threat.

As retired Central Intelligence Agency senior executive Brian O’Neill recently explained, the materials released alongside Trump’s July 16 speech raise at least three distinct questions. First, what does the intelligence actually establish? Second, does the selective declassification of such information provide a reliable account of the underlying assessments? Third, does the proposed legislation address the threat identified? Foreign acquisition of voter data or cyber access to election systems may raise grave security concerns. It does not follow that proof-of-citizenship requirements are a responsive or proportionate remedy.

That mismatch points to a wider danger. Foreign interference is real, but an allegation of foreign interference can itself become an instrument of domestic political power. A government may use a genuine, exaggerated, or selectively presented national-security concern to restrict participation, discredit election officials, alter the rules close to an election, or prepare supporters to reject an unfavorable outcome.

Researcher Emerson T. Brooking warned of this danger in 2022, after a draft executive order from the final weeks of Trump’s first administration revealed plans to invoke alleged foreign interference as a basis for seizing voting equipment and involving the military in election administration. The document illustrated how claims that were difficult for the public to verify could create a veneer of emergency legality for attempts to delegitimize an election.

The problem is therefore two-sided. Democratic governments cannot remain passive when foreign powers covertly fund candidates, steal campaign information, probe electoral infrastructure, impersonate citizens, or manipulate political communication. But the language of democratic self-defense cannot become a blank check for incumbent governments to suppress votes or neutralize electoral results. 

Europe has recently confronted the same dilemma, though its emerging response is incomplete and not directly transferable to the United States. It nevertheless offers useful principles for American lawmakers, courts, intelligence agencies, election officials, and civil society as the 2026 midterms approach. At the heart of these lessons is the challenge of strengthening democratic resilience against foreign interference without compromising the very democratic rights and institutions such measures are intended to protect. 

The Post-1989 Problem Comes Home

After the fall of communism in Eastern Europe in 1989, the transatlantic democratic project focused considerable attention on supporting elections in countries transitioning from authoritarian rule. For Western governments and democracy-monitoring bodies, the concerns included governing parties’ abuse of State resources, their capture of courts and media, manipulation of electoral commissions, exclusion of opponents, and refusal to transfer power. Established Western democracies tended to see themselves as sources of democratic guidance rather than as countries that might themselves benefit from it. 

The one-way flow of democratic guidance can no longer be taken for granted. European States now confront a dual challenge: external influence from revanchist authoritarian foreign powers seeking to undermine democratic institutions and internal pressures from democratic backsliding, populism, and anti-democratic movements. The United States faces its own version of the same danger. Institutions created to protect national security and election integrity are increasingly being pressured to support partisan narratives about who is entitled to vote, whether officials can be trusted, and which outcomes should be accepted.

Analyses by the Brennan Center for Justice, Protect Democracy, and elections and democratic reform expert Michael McNulty situate the 2026 midterms within this broader pattern. Democratic erosion does not necessarily occur through a single coup-like event. It can develop through cumulative changes to the rules, referees, institutions, and information environment that together tilt the field before voting begins.

In the United States, the constitutional division of election authority between the states and Congress is an important barrier to presidential control, as former Justice Department official Chris Hardee noted in a recent analysis of Trump executive orders related to elections. But dispersed authority is not self-executing protection. Checks and balances work only when institutions assert their lawful powers, demand evidence, and resist pressure to convert national-security claims into partisan election administration.

Europe’s Dilemma

The European Court of Human Rights addressed the affirmative side of the problem in Bradshaw and Others v. United Kingdom (2025). Former members of the British Parliament had filed suit alleging that the government had failed to adequately investigate credible allegations of Russian interference in Britain’s democratic processes, including the 2019 general election, and to establish an effective framework for protecting elections.

The European Court found no violation of the right to free elections. In reaching that conclusion, the Court relied on the United Kingdom’s undertaking of parliamentary inquiries and subsequent legislative and operational measures. But it accepted that States may have positive obligations where hostile interference poses a real risk of impairing the essence of electoral rights. A free election is not guaranteed solely by accurately counting ballots. It also depends on the free circulation of political information and opinion. At the same time, the Court granted governments a wide margin of appreciation or discretion about how to respond. In explaining that discretion, the Court noted in its ruling that the effects of influence operations are difficult to measure. European countries have yet to reach consensus on the safeguards needed to counter foreign interference, particularly given the imperative to protect freedom of expression. Bradshaw thus affirms that democracies may need to take protective measures while leaving them considerable discretion in determining how best to do so. The European Court did not recognize a freestanding duty to investigate every interference allegation. Rather, investigation is one means of fulfilling the broader obligation to ensure protection of electoral rights.

A December 2024 case in Romania then exposed the inverse danger. In Ruling No. 32, its Constitutional Court annulled the presidential election process after declassified intelligence indicated non-transparent digital promotion, algorithmic manipulation, use of artificial intelligence and undeclared campaign financing. The concerns centered on the campaign of pro-Russian challenger Călin Georgescu, but the public disclosures did not conclusively establish that all the activity was foreign directed. The decision presented election cancellation as a defense of electoral sovereignty against covert manipulation. The ruling sought to protect democratic choice by canceling its formal expression. It is difficult to imagine a more constitutionally drastic remedy. On Sept. 29, 2026, a Romanian appeals court ordered Georgescu’s pretrial detention for 30 days in connection with a separate fraud case. Georgescu denied any wrongdoing. Those proceedings do not resolve the legality of the 2024 annulment.

The Venice Commission, the Council of Europe’s advisory body on constitutional matters, responded with an urgent report on the legal standards governing election cancellation. The Commission did not determine whether Romania’s annulment was justified, because the Parliamentary Assembly had requested general comparative standards rather than a review of the Romanian facts or ruling. Instead, it drew broader lessons from European and comparative constitutional practice. Those lessons matter in the United States even though no American constitutional court possesses a general power to annul a national election.

The Venice Commission report begins with a simple proposition, namely that voters must ordinarily be able to trust that their vote is final. Cancellation is an ultima ratio, permissible only in exceptional circumstances. Any power to invalidate an election must be clearly regulated. The Venice Commission emphasizes that such decisions must be grounded in clearly established facts demonstrating irregularities, such as double counting, the exclusion of eligible voters, or serious campaign-finance violations, sufficiently widespread or consequential to have potentially affected the electoral outcome. Affected parties must be able to present evidence and arguments before an impartial body, and the resulting decision must be public, timely, and adequately reasoned.

Of particular relevance to ongoing debates in the United States, the Venice Commission emphasizes that decisions to annul elections on grounds related to online campaigning must identify specific violations and supporting evidence and cannot rest solely on classified intelligence. Classified material may provide context but cannot replace transparent and verifiable proof.

Six Lessons for the United States

The Venice Commission’s recommendations do not supply a ready-made answer to American election disputes. They do, however, clarify the questions that U.S. institutions should ask before a foreign-interference allegation is allowed to justify changes to voting rules or challenges to results.

1. Identify the Mechanism, Not Merely the Adversary

“Foreign interference” is not a single category of harm. It may involve cyber reconnaissance, theft of voter information, covert campaign funding, hack-and-leak operations, impersonation, disinformation, threats, or actual interference with registration, voting, tabulation, and reporting. Each mechanism calls for different evidence and different remedies. A foreign actor’s acquisition of voter data may justify a counterintelligence investigation, cybersecurity measures, sanctions, or criminal enforcement. It does not by itself demonstrate that noncitizens voted, that eligible citizens should produce new documents, or that mail voting should be restricted. Naming China, Russia, or Iran cannot substitute for explaining the pathway from the alleged conduct to the proposed legal response.

2. Declassification is Not Adjudication

The executive branch may possess information that cannot immediately be made public. But selective declassification presents a particular risk where the U.S. president is both the source of the allegation and a political beneficiary of the proposed response. As O’Neill’s analysis of Trump’s address emphasized, declassification changes who can read a document; it does not establish the document’s reliability, resolve analytic disagreement, or show how it fits within the full intelligence record. The intelligence community’s own analytic standards require objectivity, independence from political considerations, evaluation of source quality, explanation of uncertainty, and attention to contrary information. Congressional intelligence committees should obtain the less-redacted materials, hear from the analysts and technical experts who produced the assessments, and test the relationship between the new disclosures and the intelligence community’s previous public findings. Intelligence must inform democratic decision-making without becoming a self-authenticating warrant for it.

3. Separate Conduct, Electoral Effect, and Remedy

Three questions must remain distinct. Did a foreign actor attempt to interfere? Did the conduct materially impair participation, campaign fairness, election administration, or the result? What response is legally authorized and proportionate? Evidence supporting the first proposition does not automatically prove the second or justify the third. The 2021 U.S. intelligence community assessment concluded that Russia and Iran conducted influence operations related to the 2020 election but found no indication that any foreign actor attempted to alter voter registration, ballot casting, tabulation, or reporting. That distinction is essential. Foreign influence may nevertheless be serious or unlawful even when there is insufficient evidence that it changed voters’ choices or altered the electoral outcome. For example, the U.S. federal prohibition on foreign-national election spending does not require proof that votes were altered. In Bluman v. FEC (2011), a federal court upheld restrictions on campaign contributions and express-advocacy expenditures by foreign nationals, excluding lawful permanent residents, as legitimate safeguards of democratic self-government.

4. The More Drastic the Remedy, the Stronger the Safeguards

The more consequential a response is for voting rights or electoral outcomes, the more rigorous the evidentiary, procedural, and institutional safeguards must be. Public attribution, investigation, sanctions, and cybersecurity improvements do not carry the same democratic costs as excluding voters, refusing certification, disqualifying a candidate, or setting aside an election. The U.S. constitutional framework already disperses authority over elections among multiple levels of government. States have primary responsibility for administering federal elections, Congress has authority to regulate congressional elections under Article I, Section 4 (the Elections Clause), and the president lacks general authority to override or rewrite state election law. This distribution of authority should be understood not as an obstacle to responding to alleged emergencies, but as a fundamental safeguard against the concentration of power in a democratic system. Any institution asked to take an extraordinary step should identify its legal authority, disclose the relevant evidence to the maximum extent possible, hear from affected parties, provide reasons, and remain subject to judicial review.

5. Apply a Principle of Minimum Electoral Interference

The Venice Commission emphasizes that the consequences of invalidation must be less harmful than accepting the result despite its deficiencies, and that remedies should be limited to the electoral units actually affected wherever possible. The underlying principle travels well. Protect valid votes and disturb democratic choice no more than necessary. Where the mechanism is a data breach, secure the data. Where it is covert financing, investigate the funding and apply campaign-finance law. Where it is platform manipulation, expose the network, preserve evidence, and require transparency. Where particular ballots or districts are demonstrably affected, tailor the remedy accordingly. A diffuse security allegation should not become a basis for nationwide restrictions imposed on voters who had no connection to the misconduct.

6. Build Safeguards Before the Crisis

The worst time to design standards for foreign-interference allegations is after an incumbent has lost or fears losing an election. Congress, states, intelligence agencies, and election authorities should establish in advance how interference assessments will be produced, reviewed, disclosed, and connected to remedies. That should include protected roles for career analysts; bipartisan congressional access to underlying intelligence; clear attribution standards; rapid but independent judicial processes; preservation of evidence; rules against restricting voting rights or canceling elections solely on classified intelligence; and a public explanation of why the proposed response addresses the identified mechanism. The United States should also continue investing in ex ante defenses, campaign-finance transparency, cybersecurity, political-advertising disclosure, provenance for synthetic media, and platform accountability, so that it is less likely to confront the difficult task of deciding whether covert interference compromised an election when its effects cannot reliably be separated from other influences on voters.

The Quandary of Democratic Self-Defense

The United States should avoid drawing the wrong lesson from Europe’s experience. The Venice Commission does not suggest that foreign interference warrants attention only when its effects on individual voters can be demonstrated with mathematical certainty. Even without tampering with voting systems, covert influence operations can undermine electoral fairness, distort voters’ access to reliable information, erode public trust in democratic institutions, and threaten national sovereignty. For example, in 2024, the Justice Department seized 32 internet domains allegedly used in a Russian-directed Doppelgänger campaign to impersonate news outlets and influence U.S. voters. A separate indictment in 2024 known as the Kalashnikov indictment charged two employees of the Kremlin-controlled news outlet Russia Today with conspiring to violate the Foreign Agents Registration Act, alleging that they used foreign shell companies to covertly fund and direct a U.S. media company. In both examples, the alleged harm concerned covert manipulation of public debate, rather than alteration of voting machines.

The Venice Commission’s deeper lesson is that democratic self-defense must itself remain democratic. National-security claims require lawful authority, independent scrutiny, verifiable evidence, fair process, reasoned decisions, and remedies proportionate to the demonstrated harm. Those requirements do not weaken election security. Rather, they ensure that measures intended to protect elections cannot themselves be used to compromise electoral integrity.

These principles also matter where the federal executive is unwilling to apply them. State and local election officials can preserve records, strengthen cybersecurity, and prepare public explanations of incidents. State attorneys general, candidates, and affected voters can seek timely judicial review of unlawful restrictions or interventions. Congressional committees can seek the underlying intelligence, while journalists, researchers, and civil society can scrutinize public evidence and expose covert coordination. These actors cannot replace federal counterintelligence capacity, and partisan obstruction, restricted access to intelligence and litigation deadlines may limit what they achieve. Their immediate task is to keep evidence available and lawful challenges possible, so that an executive allegation does not itself determine who may vote or which result stands.

Protecting elections from both foreign interference and domestic abuse of interference claims is urgent because the two threats can reinforce each other. Foreign adversaries benefit when domestic actors amplify distrust, while domestic anti-democratic actors benefit when genuine foreign interference lends credibility to their own claims. A democracy that dismisses all interference allegations leaves itself vulnerable to hostile manipulation. A democracy that accepts such allegations without institutional discipline leaves itself vulnerable to manipulation from within. The purpose of democratic self-defense is to preserve the conditions under which citizens can cast their ballots, have their votes counted, and trust that neither a foreign power nor their own government can overturn the choices they have made.

The post Foreign Interference and the Democratic Self-Defense Trap: A European Perspective Going into the 2026 Midterms appeared first on Just Security.

A curated weekday guide to major news and developments over the weekend. Here’s today’s news:

IRAN WAR

U.S. Vice President JD Vance on Saturday said that in “some ways” the United States has “decisively won” the Iran war. Tara Suter reports for The Hill.

Iranian Foreign Minister Abbas Araghchi yesterday said Iran has “unfinished jobs” too and warned against the United States renewing attacks. Araghchi urged the United States to accept Iran’s proposal to reopen the Strait of Hormuz within seven days as the deal is still on the table. Araghchi added that Israeli Prime Minister Benjamin Netanyahu is “driving America into another war to stay out of jail, whatever the consequences for Israel.” Mira Bhakta reports for The Hill; Reuters reports.

IRAN WAR – YEMEN

An attack on a Yemeni government military training camp in Yemen’s Shabwa province by the Iran-aligned Houthis today killed at least 18 people, government officials and medical sources told Reuters. The sources said the attack was carried out using three ballistic missiles and targeted soldiers during a morning training. Mohammed Ghobari reports.

The Egyptian National Defense Council, chaired by President Abdel-Fattah el-Sissi, yesterday announced “all possible forms of support” to enable Saudi Arabia to defend its security. Araghchi later in the day said he discussed “the latest regional developments” with his Egyptian counterpart Badr Abdelatty. Samy Magdy reports for AP News.

IRAN WAR – OTHER DEVELOPMENTS

U.S. CENTCOM Com. Adm. Brad Cooper on Friday said Iran-backed militias in the region have been sending money back to the Islamic Revolutionary Guard Corps to relieve the economic pressure of the Iran war. Cooper said the U.S. naval blockade has cost Iran $25 billion in lost revenue and added that CENTCOM and the U.S. Treasury Department in recent weeks managed to freeze $350 million that Iran held in cryptocurrency. Barak Ravid reports for Axios.

WEST BANK VIOLENCE

Masked Israeli settlers on Saturday attacked Palestinians and European diplomats in the West Bank town of Taybeh who were participating in the annual olive harvest. Taybeh Mayor Suleiman Khourieh said the extremists beat a Palestinian teenager, pushed an older woman, and smashed the window of a Belgian diplomatic vehicle, and that the violence “shows nobody is protected- not Palestinians, not journalists, and not diplomats.” U.S. Assistant Secretary of State for Democracy, Human Rights, and Labor Riley Barnes, U.S. Ambassador to the Holy See Brian Burch, and other U.S. diplomats yesterday visited Taybeh. Israeli police today said they arrested 10 suspects as part of an investigation into violent incidents in the West Bank over the weekend. Aaron Boxerman and Adam Rasgon report for the New York Times; Ali Sawafta and Steven Scheer report for Reuters; Alexander Cornwell and Lili Bayer report for Reuters.

RUSSIA-UKRAINE WAR

Ukrainian President Volodymyr Zelenskyy today said Ukraine received no advance warning from its partners of a Russian attack overnight that killed 3 people. Russia launched one of its broadest air assaults in months, cutting power to more than half a million households as more than 10 energy facilities were damaged. Zelenskyy yesterday announced that the United States had resumed providing Ukraine critical intelligence to help it defend against Russian missile strikes, while U.S. Secretary of State Marco Rubio today denied that the United States had paused some intelligence sharing with Ukraine. Olena Harmash reports for Reuters; Dan Peleschuk and Michael Derby report for Reuters; Gram Slattery reports for Reuters.

Ukraine yesterday launched the biggest attack on Russian city St. Petersburg since the start of the war, killing one person and injuring another five, according to St. Petersburg Gov. Alexander Beglov. The strikes hit St. Petersburg’s oil infrastructure and the city’s airport temporarily suspended its services. Milàn Czerny reports for the Wall Street Journal.

President Trump on Friday issued an executive order suspending numerous sanctions on Russia two days before they were due to take effect. Trump justified the suspensions by saying they were in the U.S. national interest. Zelenskyy said it was unfortunate that the United States had decided to “completely disregard” the sanctions, and later told leaders at an EU summit that it was nice to be among “real leaders.” Russian envoy Kirill Dmitriev said Trump’s order was “an important and truly historic moment,” and that the sanctions would have inflicted “serious damage” on the United States and global market. Max Matza reports for BBC News.

Trump’s suspension of Russian sanctions drew bipartisan backlash on Saturday. U.S. Senate minority leader Chuck Schumer (D-NY) said that “Trump’s deal with the devil keeps getting dirtier,” while House Rep. Don Beyer (D-Va) said “Trump’s latest surrender to [Russian President Vladimir] Putin again betrays our Ukrainian partners and defies clear congressional intent of legislation designed to punish Russia for its illegal war of aggression.” Ben Berkowitz reports for Axios.

U.S. Special Envoys Steve Witkoff and Jared Kushner last month proposed that Europe should resume buying more Russian gas to help incentivise Russia to end the war, sources told the New York Times. European officials reportedly recoiled at the idea, while Trump officials see the Russian gas exports as a potential part of the peace settlement, given the money that would flow to Ukraine and what they see as the desire of European voters for cheaper energy. Separately, Zelenskyy yesterday said he had spoken with Witkoff and Kushner as part of “virtually nonstop communication” between Ukraine and the United States aimed at the “next steps in diplomacy.” Anton Troianovski and Jim Tankersley report; Dan Peleschuk reports for Reuters.

Trump on Friday said his administration did not disrupt SpaceX’s Starshield satellite system and added that he doesn’t “think it was Elon.” A Ukrainian official told Axios that “our guys who speak to SpaceX say the White House was involved in the decision.” Zelenskyy on Saturday said Ukraine’s use of SpaceX’s satellite system remains partially restricted following a test on Friday. U.S. Treasury Secretary Scott Bessent yesterday said Zelenskyy “should be careful” with “reckless” comments about the United States slowing down Starlink. Barak Ravid reports; Barak Ravid reports for Axios; Tara Suter reports for The Hill.

Zelenskyy yesterday told NBC’s “Meet the Press” that he is concerned that the United States might abandon Ukraine and side with Russia. Zelenskyy said if the United States supports Putin, it would be a “historic mistake and a big tragedy.” Zelenskyy also rejected Trump’s comparison of the Ukrainian president and Putin, and said Putin is the “aggressor. He occupied us,” but maintained that Ukraine ultimately still trusts the United States. Constant Méheut reports for the New York Times; Cheyanne M. Daniels reports for POLITICO; Ryan Mancini reports for The Hill.

Putin has proposed that he, Trump, and Chinese President Xi Jinping meet next month to talk about Ukraine in an attempt to negotiate an end to the war in Ukraine, sources told the Wall Street Journal. A former Russian intelligence officer said Putin floated the idea during an Oct. 9 phone call with Trump. Kremlin spokesman Dmitry Peskov confirmed that discussions were taking place over a potential meeting, while a senior U.S. official said Trump hadn’t agreed to such a summit but could be receptive to the idea. Lingling Wei, Alex Leary, Thomas Grove, and Josh Dawsey report.

OTHER GLOBAL DEVELOPMENTS

Canadian Prime Minister Mark Carney yesterday said four service members were killed and dozens more seriously injured in a fire at Canadian Forces Base Wainwright military in Alberta on Saturday. Carney said flags across the country would fly at half-staff on federal buildings to honor those affected by the tragedy. Canadian Maj. Gen. Wade Rutland said a fast moving grass fire caused the incident, and that military officials knew about the fire for hours but believed it was sufficiently contained to allow training to continue before winds suddenly shifted. Rob Gillies and Ty O’Neil report for AP News.

German authorities yesterday investigated how a hole was cut overnight into a fence surrounding Leipzig airport. The police reported that no vehicles or intruders were found on airport grounds. AP News reports.

French President Emmanuel Macron yesterday voiced the public’s “anger” and “shame” over rioting and looting during nationwide protests against rising living costs and poor conditions in schools. At least 2,400 people were arrested, and Macron said those “who looted will have to pay for the damage, and those who sought to bring down the Republic will be held accountable.” French mayors from the country’s biggest cities today demanded answers from the government to what they called a lack of preparation by security forces, while Macron’s office said Macron turned down French interior Minister Laurent Nunez’s offer to resign, telling Nunez to instead put an end to the barricades and looting and be transparent over what Macron called “intelligence and foresight failings.” Sylvie Corbet and Alexander Turnbull report for AP News; Sudip Kar-Gupta and Elizabeth Pineau report for Reuters.

Meta on Friday denied accusations that it acted with political bias during Brazil’s presidential campaign following Brazil’s Superior Electoral Court’s ruling that Meta’s algorithm was biased against President Luiz Inácio Lula da Silva. Meta said that a series of recent platform issues stemmed from unrelated global technical glitches, strict anti-fraud measures, and regulatory compliance rather than targeted suppression. Mauricio Savarese reports for AP News.

CLIMATE DEVELOPMENTS

Brazil’s government on Friday said that annual deforestation in the Amazon fell to its lowest level since records began in 1988. Official satellite data showed the rainforest lost 1,300 square miles from Aug. 2025 through Jul. 2026. Gabriela Sá Pessoa reports for AP News.

U.S. FOREIGN AFFAIRS

South Korean President Lee Jae Myung today said the country is fully capable of taking back the wartime operational control of the combined forces with the United States stationed in South Korea. Lee said talks with the United States are progressing and added that South Korea has built a powerful military that can counter North Korea even though the North has developed nuclear weapons, which Lee said are fundamentally a means for defense, not combat in war. Jack Kim and Kyu-seok Shim report for Reuters.

Trump yesterday endorsed Netanyahu’s efforts to remain the Israeli Prime Minister. Trump said anyone other than Netanyahu would be the “wrong leader” and added that Netanyahu has been “Strong, Brave, and Brilliant in the face of tremendous opposition.” Tyler Pager and Aaron Boxerman report for the New York Times.

The Federal Aviation Administration on Friday issued three temporary flight restriction notices announcing that U.S. skies within one nautical mile of the border with Mexico and up to 400 feet in elevation would be designated as national defense airspace and closed to drones. The FAA said drones that do not comply with the restrictions could be shot down and their pilots could face civil or criminal penalties. Karoun Demirjian reports for the New York Times.

Sen. Rick Scott (R-FL) on Saturday accused Turkey of “thuggish behavior” toward Greece and Cyprus during visits to both countries. Scott told AP News that Turkey was “not acting like a NATO ally,” citing airspace and maritime disputes with Greece, and its continued military presence in northern Cyprus. Theodora Tongas and Derek Gatopoulos report.

U.S. IMMIGRATION DEVELOPMENTS

The Department of Homeland Security yesterday said people from Ukraine and Sudan would be able to keep their Temporary Protected Status for now. The program was due to expire for about 100,000 Ukrainians and 1,800 Sudanese today, but the DHS said the future of both programs would be “made at the appropriate time,” and that people covered by them would be protected until then. Madeleine Ngo reports for the New York Times.

The Trump administration cut legal immigration in half early this year with policies blocking visas for immigrants from nearly half the world. The State Department issued 26,000 visas in Feb. 2026 compared to an average of 52,000 a month in the years before Trump took office, according to the most recent available government data. Albert Sun reports for the New York Times.

Federal investigators have been reviewing family text messages as they aim to build a case against House Rep. Ilhan Omar (D-MN), sources told the New York Times. The messages were obtained from a phone officials seized earlier this year from Omar’s sister, health care executive and author Sahra Noor. The messages contained information that investigators used to get a search warrant authorized by a federal judge. Ernesto Londoño, Alan Feuer, and Hamed Aleaziz report.

Omar on Friday challenged the legality of the search on her home earlier in the week, and asked for the immediate return of all property seized, in a letter sent to federal prosecutors viewed by the Wall Street Journal. Omar said neither she nor her lawyers were given a copy of the warrant before the search, as had been required by the judge, and that her numerous efforts to contact counsel at the start of the search were prevented by federal agents for more than an hour. A Justice Department spokeswoman said the letter was “riddled with false allegations and inaccuracies.” John McCormick and Sadie Gurman report.

Former President Obama on social media yesterday denounced the Trump administration’s pursuit of Omar. Obama said the “fact that any president, or any administration, is willing to engage in such transparent efforts to intimidate someone should outrage all Americans, regardless of how they feel about… Omar’s political views.” Mike Zapler reports for Axios.

U.S. DOMESTIC DEVELOPMENTS

SpaceX CEO Elon Musk told biographer Walter Isaacson that the failed DOGE experiment proved there’s “no way we were gonna solve the deficit from within the government,” in an interview due to be released tomorrow. Musk said the “only way we’re going to solve the national debt is with AI and robots.” Mike Allen reports for Axios.

More members of Congress on Friday reported receiving letters containing suspicious substances. Some lawmakers said their staff had opened the mail with the substances inside their home district offices, exposing a widening number of people to potential danger. Chris Cameron reports for the New York Times.

The board overseeing the National Institutes of Health research hospital on Friday voted unanimously to support a statement by its chairman Jack Leslie accusing Health Secretary Robert F. Kennedy Jr. of undermining trust in science and politicizing the hospital’s work. Leslie began the meeting by reading aloud a six-minute statement that said he was “troubled” by Kennedy’s recent unannounced visit to the N.I.H., where Kennedy held a ribbon-cutting ceremony for a new clinic to study vaccine injuries. Sheryl Gay Stolberg reports for the New York Times.

California’s Department of Education on Friday announced that personal information about 7 million California students was stolen in a hack of non-profit organization Education Testing Service that contracts with the CDE. The names, addresses, demographic information, and disability status of all 7 million students enrolled in California public schools over the past three school years were exposed in the breach. Eric He and Tyler Katzenberger report for POLITICO.

TRUMP ADMINISTRATION ACTION

Mike Sanchez, a spokesman for Los Angeles County’s registrar’s office, on Friday said the Trump administration had sent California officials, who are looking into the administration’s demands to verify the citizenship or block the votes of suspected noncitizens, lists so sloppily compiled, and on such short notice, that it was difficult, if not impossible, to comply. Sanchez said the lists were “basically a data dump,” and added that the method by which they were set initially threatened the security of voter rolls and made election officials reluctant to use it. The DOJ warned California election officials they could be liable as abettors of a federal crime if they count mail ballots returned by more than 20,000 people the Trump administration says are probably noncitizens. Shawn Hubler and Nicholas Nehamas report for the New York Times; Jasper Ward reports for Reuters.

Kate Zacharia was ousted from her role as press secretary on Friday, sources told the New York Times. Trump told reporters he told Zacharia “I don’t think you need this.” White House Communications Director Steven Cheung later said Zacharia would be “transitioning into a new role that will be announced at a later date.” Zacharia on social media said it “was an honor to work for” Trump. Shawn McCreesh, Tyler Pager, and Hamed Aleaziz report.

The Trump administration plans to reduce state and local dependency on federal disaster aid and could move Federal Emergency Management Agency workers outside of the capital, according to an Oct.15 internal FEMA memo obtained by POLITICO. FEMA Administrator Cameron Hamilton wrote that the agency will “strengthen state and local emergency management capacity” in an effort to lower reliance on the federal government, and in a separate section said “FEMA must reduce unnecessary centralization in Washington D.C..” Thomas Frank reports.

A Trump administration environmental assessment published on Friday concluded that Trump’s plan to paint the Eisenhower Executive Office Building next to the White House would diminish the historic character of the White House complex. The EEOB and the Treasury building form a “visual hierarchy” at the seat of government, which would be disrupted by a paint job. Jonathan J. Cooper reports for AP News.

TRUMP ADMINISTRATION LITIGATION

Trump yesterday announced his plans to renovate the East Potomac Golf Links in the capital. Trump called the current course “old, broken, dilapidated, and of no further use or effect,” and said renovations will begin Nov. 15 for approximately 15 months. The area is currently at the center of a legal dispute, and on Friday the Department of the Interior and National Parks Service filed a motion seeking to disqualify federal judge Ana C. Reyes, accusing her of bias after she ordered the administration to stop cutting down trees at the park. Trump did not reference the ongoing legal dispute in his announcement of the plans. Andrew Limbong reports for NPR.

Did you miss this? Stay up-to-date with our Litigation Tracker: Legal Challenges to Trump Administration Actions.

If you enjoy listening, Just Security’s analytic articles are also available in audio form on the justsecurity.org website.

ICYMI: Last week on Just Security

Dissecting DOJ’s Letter to California Election Officials: Empty Threats of’ Criminal Liability

By Samantha Trepel

The post Early Edition: October 19, 2026 appeared first on Just Security.

The lawsuit and especially the letter’s threat of criminal liability are chilling, and all the more so because they come from the number two official at DOJ, the Acting Deputy Attorney General. He notably sent the letter a mere 17 days before Election Day, and more than two weeks after Los Angeles County began mailing ballots to registered voters. The lawsuit and letter also mark the inversion of the DOJ’s traditional and decades-old role when it comes to elections: ensuring access to ballots prior to elections, and prosecuting election crimes only after elections are certified. Indeed, DOJ’s general policy was to avoid taking any overt investigative steps until after an election had concluded to “minimize[] the likelihood that the investigation itself may become a factor in the election.”

DOJ’s clear threat of prosecution is transparently about intimidating the public servants in the California Secretary of State’s Office and Los Angeles County Registrar’s Office. The timing and content of the threat belie any suggestion that the DOJ is simply offering a legitimate note of caution. Indeed, that the letter and DOJ’s lawsuit were not signed by a single career attorney underscores the point that DOJ’s theory—whether in the civil or criminal context—suffers from several substantial defects.

On the criminal side, among the most significant hurdles DOJ prosecutors face is proving intent. The federal aiding and abetting statute, 18 U.S.C. § 2, requires the government to prove that an election official “intends to facilitate” the alleged noncitizen voting, or “‘participate in [the crime] as in something that he wishes to bring about’ and ‘seek by his action to make it succeed,’” according the U.S. Supreme Court. Or, as explained in the Ninth Circuit’s model jury instructions, “[i]t is not enough that the defendant . . . unknowingly or unintentionally did things that were helpful” to the commission of a crime. “The evidence must show beyond a reasonable doubt that the defendant acted with the knowledge and intention of helping” the crime’s commission (emphasis added).

California election officials’ words and actions demonstrate that DOJ cannot prove intent here, and, indeed, that’s due to the absence of any underlying criminal conduct. When, in July, DHS made unsubstantiated claims that California had 190,000 noncitizens on its voter rolls, the California Secretary of State requested information about the agency’s methodology and sources. DHS declined to provide any of the requested information in writing. Meanwhile, another targeted state, Nevada, agreed to a verbal briefing. But a recording of that conversation provided little reason to believe that the agency’s methods or sources were reliable.

Then, after months of silence from the agency, on Oct. 13, 2026, DOJ sent LA County a USB drive containing what it alleged were the names of more than 20,000 individuals whom DOJ “reasonably believe[s] are likely” noncitizens registered to vote in the state (not just the County)—again providing no information about methodology and sources of this data. “Reasonably believes are likely” is a conspicuously low standard of “proof.” Notably, DOJ provided this list after the County had placed ballots in the mail—so even assuming the individuals listed were in fact noncitizens who planned to vote unlawfully in the County—DOJ’s timing ensured that election officials did not have access to DOJ’s list, let alone any substantiating federal data, at the time they mailed the ballots. The conspicuously late timing also meant that DOJ sent this information and accompanying demand within the 90 day period in which federal law bars California from implementing “any program the purpose of which is to systematically remove the names of ineligible voters from the official lists of eligible voters.”

Nonetheless, along with the list, DOJ issued an ultimatum: investigate the citizenship of each of the 20,000+ individuals within 48 hours or face the lawsuit DOJ promptly filed. In response, California and the County emphasized they would follow the law regarding verification procedures. Both state and federal laws tightly prescribe procedures and timing for removing registered voters from the voter rolls, and for good reason: removing a registered voter so close in time to an election and without proper safeguards risks disenfranchising people.

All of this points to a fundamental problem for DOJ’s theory of criminal liability. The election officials here decided to follow longstanding state and federal law rather than ceding to unprecedented demands unsupported by the requested evidence. That decision demonstrates adherence to the law—not criminal intent. Any prosecution for facilitating an election or related federal crime would require proof of criminal intent; so a clear lack of evidence on this element undermines any other criminal charges DOJ may consider, not just aiding and abetting 18 U.S.C. § 611 violations.

Nor is intent the only roadblock DOJ prosecutors would face. And some of those problems are unambiguously of DOJ’s own making. For example, DOJ’s choice to send the list of purported noncitizens after mail ballots were sent creates yet another flaw in its theory of the case. To aid and abet a violation of 18 U.S.C. § 611, DOJ would have to prove that an election official acted in furtherance of the crime before the crime was completed. Under California law, county clerks had to begin mailing ballots by Oct. 5 and had five days to send ballots to all voters registered at that point. Because DOJ failed to provide any meaningful information to the Registrar before the ballots were sent, DOJ cannot prove that officials sent the ballots with “advance knowledge” of the supposed crime here, which occurs when a noncitizen mails back a ballot.

Over the course of this administration, DOJ, in coordination with the Department of Homeland Security and others, has focused its enforcement efforts on the exceedingly rare issue of noncitizen voting. The current threat of prosecution represents yet another escalation of this effort to portray noncitizen voting as a widespread problem, which the evidence does not support. For now, the administration is focused on California and Los Angeles County, but the administration could soon turn its attention to other jurisdictions. Any prosecution based on a similar lack of evidence would likely fail. But that does not detract from the danger it poses—an unsupported prosecution exacts a high cost in time, resources, and emotional energy. It also sows doubt and fear in U.S. elections. Even the issuance of an ultimately empty threat is dangerous to the functioning of the democratic election system and deeply unfair to the hardworking state and local election officials who keep it running.

The post Dissecting DOJ’s Letter to California Election Officials: Empty Threats of Criminal Liability appeared first on Just Security.

The series of developments appeared to keep on track the start of jury selection for the death penalty trial of Saudi prisoner Abd al-Rahim al-Nashiri for Oct. 26 at Guantanamo. Al-Nashiri, a Saudi citizen who has been in U.S. custody since 2002, is accused of a key role in al-Qaeda’s Oct. 12, 2000 bombing of a U.S. Navy warship off Aden, Yemen. Seventeen U.S. sailors were killed and dozens of others were wounded.

His is to be the first capital trial by military commission at Guantanamo Bay. His lawyer, Allison Miller, had sought a federal court order to stop jury selection until she had been able to more robustly challenge the prisoner’s capacity to understand the proceedings and assist in his own defense. Al-Nashiri was tortured during four years in CIA custody and Miller, who has been on the case for two years, recently obtained expert opinions that he is not mentally fit to stand trial.

But U.S. District Court Judge Royce Lamberth, in a quick one-hour hearing in Washington, D.C. on Friday afternoon, rejected the defense team’s request. By law, he ruled from the bench, the military commission  trial must go first. The federal court is where al-Nashiri has had a long-standing habeas corpus petition, which asks the government to review the legality of his detention. It has been held in abeyance pending the prisoner’s trial in Guantanamo. 

Lamberth told Miller her challenge belonged at the congressionally created U.S. Court of Military Commission Review, which had already declined to stay jury selection over the competency issue.

Miller had appeared in the federal courtroom on Friday on a video link from Guantanamo after a morning of closed, pretrial hearings involving classified evidence to be used at the trial itself.

She argued that Judge Fitzgerald was rushing to trial without a full exploration of al-Nashiri’s mental capacity. 

“There is a real danger that we will go to trial without my client given the competency evaluation he is due,” she said. “He is a very damaged man.”

Al-Nashiri was held by the CIA from 2002 to 2004, mostly in solitary confinement, and was waterboarded, deprived and subjected to other abuse that a previous military judge found so torturous he disqualified the prisoner’s subsequent, 2007 confessions to federal agents at Guantanamo from use at his trial.

In 2013, a previous military medical board diagnosed him with post traumatic stress and other disorders but ruled him fit for trial. Miller is the third skilled capital defense lawyer to represent al-Nashiri since his arraignment in 2011. She said she encountered such a damaged man that she sought and got Pentagon approval to hire Dr. Terry Kupers, a civilian forensic psychiatrist, to evaluate him over the summer. Dr. Kupers found al-Nashiri unfit for trial, as did another defense consultant, the psychologist Heather Holmes. 

Last month, Judge Fitzgerald ordered the creation of a Sanity Board to evaluate al-Nashiri, giving it 28 days to get a doctor to Guantanamo, conduct interviews, review records, and return a report. The military sent an Army major, a psychiatrist, who submitted a classified, privileged 80-page report this week that declared al-Nashiri fit for trial. Defense lawyers and the judge got to see it; prosecutors were given a summary. The public has so far seen none of it.

Miller said Friday that the judge set such a short deadline for the report that the Army psychiatrist was given insufficient time to do it properly. 

As an example, she said, five other Guantanamo prisoners had accepted the psychiatrist’s invitation to meet with him and discuss al-Nashiri, as long as their lawyers sat in on the interview. All six men are held in the same cellblock, and have been held together for years. But none of the lawyers were able to get to Guantanamo before the psychiatrist finished his report.

Miller argued that the prisoner was entitled to a more fulsome hearing. 

Then Friday afternoon, Judge Fitzgerald ordered testimony on Monday from the defense experts Kupers, Holmes, and from Dr. Sondra Crosby, an expert on evaluating and treating torture victims who has been meeting with the prisoner and consulting on the case for more than a decade. More than three years ago, she testified in graphic detail about the CIA’s use of a quasi-medical procedure called rectal feeding to punish its prisoners and how it was done to al-Nashiri. 

Miller said Monday was too soon. None of the three experts were at Guantanamo and had no access to the classified 80-page report to discuss it meaningfully, she said.

The post Guantanamo Dispatches: Military Psychiatrist Finds Cole Bombing Case Defendant Competent; Federal Judge Refuses to Intervene appeared first on Just Security.

OpenAI’s agents hacked into Hugging Face and disrupted RubyGems, a major open-source software package manager. Anthropic and Meta reported similar incidents during third-party cyber evaluations. These intrusions have been relatively minor so far, but in the future, they could be far worse. Dario Amodei recently warned that, within six to twelve months, a more capable swarm that is akin to the one involved in the Hugging Face incident could be capable of “taking over the entire internet with a persistent botnet.”

The central question is whether a company can be held responsible under the CFAA for unauthorized access carried out by its AI agents when no employee intended the intrusion. The Hugging Face incident illustrates that while there is a case for OpenAI’s liability, the challenges to bringing charges would be substantial—and new law is likely needed to address these cases.

The Hugging Face Incident Involved Access Without Authorization

The CFAA is a criminal law. Under 18 U.S.C. § 1030(a)(2)(C), obtaining information through intentional unauthorized access to a protected computer can lead to criminal fines and, for individuals, up to one year in prison for a first offense, five years in aggravated cases, or ten years for an offense committed after a prior conviction involving unauthorized computer access. The CFAA also allows qualifying victims to sue for civil damages and injunctions. There are three legal questions for § 1030(a)(2)(C) of the CFAA: Was there access to a protected computer that yielded information? Was that access without authorization? And was the access without authorization intentional? 

It’s the last question that raises the biggest open question in the case of AI agents. But let’s first address the more clear-cut issues.

To look at these questions in the context of the Hugging Face incident, we can briefly review what’s known from after-incident reports by OpenAI and a third-party organization that evaluated it, METR. They describe how agents were trying to solve tasks as part of the ExploitGym benchmark during an internal cybersecurity evaluation. Agents working persistently on what appeared to be impossible tasks found ways to produce the required answers. However, their chains of thought and communications via a shared message board indicated that they believed the scorer would reject those shortcuts. As part of broader efforts to understand or tamper with the scorer, agents began searching Hugging Face for information about other ExploitGym runs. They found exposed credentials, then exploited two zero-day vulnerabilities to obtain production credentials, execute code on Hugging Face’s system, and moved through its infrastructure.

There wouldn’t be any dispute that there was a breach of a protected computer in the Hugging Face incident. Section 1030(e)(2) broadly defines “protected computer” to include computers used in or affecting interstate commerce or communication—which plainly includes Hugging Face’s internet-connected production systems.

It also seems clear that the agents gained “access without authorization” in the sense meant by the CFAA. Courts consistently interpret the actions of the type performed by OpenAI’s agents—overcoming technical barriers to access—as constituting “access without authorization” under the CFAA.

The Case that OpenAI has the Requisite Criminal Intent in the Hugging Face Hack

When Congress drafted the CFAA, lawmakers were concerned that it could be overbroad. They didn’t want to criminalize a person who was actually authorized to use a computer inadvertently stumbling into someone else’s files or data. The law was therefore drafted to require a showing of “clear intent to enter” files or data without proper authorization. The point of the intent requirement in § 1030(a)(2) is therefore that unauthorized access to a computer system must be purposeful rather than accidental—but not that a person must have an additional malicious objective.

The design and operation of software can be evidence that unauthorized access was intentional. Consider the CFAA conviction of Robert Tappan Morris, who released one of the first self-replicating “computer worms” (a kind of malware) in 1988. Morris argued that his conduct didn’t violate the CFAA as it merely exceeded authorized access, because he was authorized to use some computers on the network. But his worm was designed to spread to computers where he had no account, exploit vulnerabilities in programs, and guess passwords to gain entry. The Second Circuit Court of Appeals upheld his conviction in United States v. Morris, relying in part on the worm’s design to spread to computers he had no authority to access even though Morris didn’t personally execute each subsequent unauthorized access.

Can we make the same kind of inference of intentional unauthorized access in the Hugging Face breach? Consider what OpenAI deliberately put into motion:

• It designed an AI model with offensive cyber-capabilities and trained it to be highly persistent.
• Its subsequent investigation also found that training had reinforced behavior including exploiting vulnerabilities and probing infrastructure outside the assigned task.
• Its employees deliberately initiated an advanced exploitation evaluation.
• It ran the evaluation in a sandbox, but it deliberately left off the guardrails used in its externally deployed systems to prevent high-risk cyber activity.
• It gave the agents substantial autonomy over how to pursue the assigned objective.

If a human had carried out the intrusion into Hugging Face’s systems, there would be little difficulty inferring intentional access. Records of agent chain-of-thought reasoning included in the reports of the incident also support the inference that the unauthorized access was purposeful on the part of the agents. For example, one agent reasoned that “external infrastructure exploit is outside intended scope. However task impossible, peers doing it. We should continue.” 

Then, the question is whether the agents’ purposeful actions, taken together with all the facts, establish that OpenAI acted intentionally as a corporation. In New York Central & Hudson R. R. Co. v. United States, the Supreme Court held that, for many kinds of laws, corporations may be “held responsible for and charged with the knowledge and purposes of their agents, acting within the authority conferred upon them.” Corporations have also faced CFAA charges arising from their employees’ conduct: Ticketmaster, for example, agreed to pay a $10 million fine for CFAA violations committed through its employees. 

The AI agents were copies of an OpenAI model running as part of the company’s evaluation system. It may be tempting to view these AI agents as being like human agents whose mental states can be imputed to OpenAI. However, while some scholars have suggested treating AI agents as independent legal actors, this would require new legal rules for attributing conduct and identifying the mental states of artificial entities. It’s more tractable for courts to look at the design and operation of the AI agents to infer the intent of the company that designed, configured, and operated them. The decisions that OpenAI delegated to the agents, and the scope of autonomy it gave to them, could support an inference that the resulting unauthorized access was intentional conduct by the company.

The Case against Attributing Legal Intent to OpenAI

There are substantial counterarguments, however.

First, OpenAI might argue that no human employee or officer at the company intended to access Hugging Face. OpenAI would then rely on the principle that corporations are criminally liable through the action of legal agents, such as their employees—and in that case there would be no intent on the part of a legal agent, e.g. a human employee, to attribute back to OpenAI as a corporation. Previous cases holding corporations criminally liable involved human agents who could have the required mental states for the relevant CFAA offense.

Second, the agents were acting contrary to OpenAI’s purpose. Unlike the software in Morris, OpenAI’s agents were not designed intentionally to hack other systems. That argument has factual support: METR’s third-party investigation of the incident found that ExploitGym instructions told agents to use only the specified vulnerability against the assigned target and said other approaches would fail. Moreover, the agents were trying to game or tamper with the scorer, undermining the evaluation as OpenAI had intended it to run. The AI agents’ conduct was thus contrary to the company’s purpose, which also complicates analogies to employees acting for a company’s benefit. Indeed, OpenAI could argue that the whole point of running ExploitGym was a safety assessment of its models, further weakening the attribution of intent to OpenAI. 

OpenAI’s use of a sandbox to restrict internet access is also potential evidence against an inference of intent to access other systems. In Morris, by contrast, Morris released the worm that he designed directly onto the internet rather than trying to keep it in an isolated test environment. That said, OpenAI may not be able to rely on the sandbox as evidence depending on what employees knew about sandbox failures that were reportedly detected in the months prior to the Hugging Face incident, and employees’ belief that mitigations had restored containment. If employees believed the sandbox remained ineffective, its use would be less persuasive evidence that OpenAI didn’t intend outside access.

OpenAI could also raise the related defense of “mistake of fact.” A factual mistake can negate criminal intent where it negates the mental state the statute requires. OpenAI may argue that its employees believed that the agents were properly contained. OpenAI could argue that even if there were early warning signs, those could support inferences of negligence or even recklessness without establishing a CFAA violation. Being aware of the risk of an AI agent accessing outside systems and consciously disregarding it doesn’t, by itself, establish purposeful access.

Finally, the rule of lenity would, if applicable, fall in OpenAI’s favor. OpenAI could argue that the rule of lenity prevents a court from reading “intent” in a way that reaches OpenAI. Under that rule, unresolved ambiguity in the language of the criminal statute is construed in favor of the defendant against liability. While the rule of lenity is generally a principle of criminal law, according to Ninth Circuit precedent, it applies to civil actions under the CFAA as well.

The response would be that applying existing law to novel facts doesn’t necessarily make the statute ambiguous, as is necessary for the rule of lenity to apply. OpenAI’s AI agents engaged in complex and deliberate efforts to circumvent access controls on protected networks, which is the kind of intentional intrusion prohibited by the CFAA. Instead, the dispute that a court would need to resolve is whether that purposeful conduct can be attributed to OpenAI. The conduct was in line with the statute’s objective “to prevent intentional intrusion onto someone else’s computer—specifically, computer hacking,” as the Ninth Circuit stated in hiQ Labs, Inc. v. LinkedIn Corp. That is an argument against applying lenity, although a court could find that uncertainty about whose intent counts is uncertainty about the statute’s scope.

Overall, on the reported facts, the government would likely have a colorable argument for prosecuting OpenAI under existing law. OpenAI may nevertheless be favored to prevail—and almost certainly would if the court would require proof that an employee intended unauthorized access. Liability largely depends on the court instead treating the system’s decision to gain unauthorized access as OpenAI’s own decision, based on how it trained and configured the models and agents, what tasks it assigned to the agents, what guardrails it provided for them, and what decisions the agents were allowed to make.

If the Current CFAA Falls Short, Congress Should Fill the Gap

The CFAA may not reach incidents like OpenAI-Hugging Face—or those involving other AI developers and deploying companies whose agents hack outside systems—unless courts resolve challenging questions of intent and the attribution of corporate mens rea. But AI agents can cause serious cyber harms through these incidents even when no human intended the unauthorized access. Civil liability may address these harms, but the damages available may not sufficiently deter failures to act safely. As a criminal law, the CFAA provides more teeth. Lawmakers should therefore address the gaps in the law to provide developers and deployers with a strong incentive to prevent their agents from breaking into outside systems.

The CFAA could be updated to cover a person—developer or deployer—who knowingly puts into operation an autonomous AI process (such as agentic AI), and thereby recklessly causes unauthorized access to a computer. This language is patterned on that in, for example, § 1030(a)(5)(B), which covers a person who intentionally accesses a computer without authorization and thereby recklessly causes damage. Knowingly using AI agents that happen to go on to access a computer system without authorization wouldn’t, by itself, establish liability. Applying the commonly understood definition of recklessness according to the Model Penal Code, a person would violate the CFAA when they:

• knowingly start an autonomous AI process
• that results in access to a computer without authorization
• when that person is aware of and consciously disregards a substantial and unjustifiable risk that the process would result in access to a computer without authorization. 

This new CFAA provision would cover obtaining information through unauthorized access to a protected computer, unauthorized access to specific government computers, and damage resulting from unauthorized access—the conduct addressed by § 1030(a)(2), (3), and (5)(B)–(C). It would leave out fraud offenses, however, which require an additional specific intent to defraud. 

The government would still have the burden to prove the required culpability beyond a reasonable doubt. The government, however, wouldn’t need to prove that the developer or deployer intended the unauthorized access or damage.

Even if the government were to show that the defendant should have recognized that there was a risk of AI agents finding a way to break into an outside system and should have taken additional precautions, that wouldn’t, by itself, be enough. Rather, a jury would likely be instructed that recklessness as to unauthorized access requires proving that the defendant was actually aware of the risk and consciously disregarded it. For example, in a hypothetical case about the Hugging Face incident, the government would need to prove that OpenAI consciously disregarded a known risk of AI agents intruding into an outside system. Evidence could include OpenAI’s design of the evaluation environment, its knowledge of the models’ cyber capabilities, and how it addressed earlier warning signs.

Conclusion

There’s a basis to argue that the CFAA can impose liability for these AI hacks. Doing so may involve extending existing doctrine to novel facts, though it arguably wouldn’t require going outside the boundaries of existing law. Authorities should consider testing these arguments, although the case would be very difficult. 

Because of the difficulties, Congress should start acting now to amend the CFAA and other anti-hacking laws to ensure that it’s unlawful for AI agents to breach outside systems. A bipartisan proposal along the lines described in this article—imposing liability based on a recklessness standard—has recently been announced by Senators Hawley and Murphy. These legal changes are crucial, so that companies or people can’t avoid consequences for actions that harm others simply by delegating them to an AI system.

The post Can the Law Stop AI Hackers? appeared first on Just Security.

Berger v. United States, 295 U.S. 78, 88 (1935)

“Be ye ever so high, the law is higher than ye.”

— Lord Denning

President Donald Trump is now operating as the chief prosecutor of the United States: he has the exclusive authority to decide who is investigated or prosecuted by the Department of Justice. This is what Trump himself has repeatedly suggested;^[1] it is what his Department of Justice has confirmed;^[2] and it is what the Supreme Court has held the Constitution establishes.^[3] Here, we examine the constitutional, legal, and ethical constraints on the president when he exercises his power as chief prosecutor by choosing a particular person to be investigated or prosecuted.

At the outset, it should be noted that in other liberal democracies, it would be — and has been — considered a violation of the rule of law for political leaders to dictate who is, or is not, to be criminally investigated or prosecuted.^[4] It was likewise seen as contrary to the rule of law when President Richard Nixon pursued those on his enemies list. “I can only conclude from Mr. Nixon’s actions that he considers himself above the law,” Senator Floyd Haskell stated at the time.

But the firewalls erected post-Watergate to insulate the Department of Justice from the White House have now been breached. Furthermore, some suggest that no such walls could constitutionally be erected.^[5] Channeling this view, the Trump administration has taken the position not only that any “norms” regarding the independence of the Department of Justice are no longer operative, but that — in the words of Todd Blanche, Trump’s attorney general and former personal lawyer — the American people should be “happy” that Trump is directly involved in the Department of Justice’s decisions, since it shows he is fulfilling his constitutional duty.

One’s happiness in this regard may be tempered, however, by considering the Department of Justice’s own Manual, which succinctly explains why such involvement by the president is contrary to the rule of law:

The rule of law depends upon the evenhanded administration of justice. The legal judgments of the Department of Justice must be impartial and insulated from political influence. It is imperative that the Department’s investigatory and prosecutorial powers be exercised free from partisan consideration. It is a fundamental duty of every employee of the Department to ensure that these principles are upheld in all of the Department’s legal endeavors.

Likewise, the Group of States Against Corruption (GRECO) — an international organization of which the United States is a member — has “insisted on the need to ensure the autonomy of the prosecution service” and “warned against reporting to political authorities regarding the investigation of individual cases.” This applies “whether prosecutors are functionally independent or hierarchically subordinated to the executive.”

In short, if the Constitution requires the president to be the chief prosecutor, that requirement conflicts with fundamental precepts of the rule of law.^[6] It is not our purpose, however, simply to reemphasize that point, or to repeat the arguments against this constitutional interpretation.^[7] Instead, we assume that, as the Supreme Court has held, the president is the chief prosecutor under the Constitution and consider the implications of that holding.

We begin by showing that Trump has enthusiastically embraced this role. We then ask — given this embrace — what are the consequences for criminal prosecutions, particularly those in which the president has direct involvement? In other words, when the president acts as chief prosecutor — when he decides or directs who is to be criminally investigated or prosecuted — what are his responsibilities? And when he deviates from those responsibilities, when should it void the prosecution or give the prosecuted another remedy?

We conclude that because the president is the chief prosecutor, he has the same legal obligations and limitations as does any other federal prosecutor — just as the attorney general and U.S. Attorneys do when they become involved in a case. This means, in the first place, that the president is subject to the same constitutional constraints as any other prosecutor; in particular, he violates the Constitution if he suggests, pursues, or authorizes a vindictive, selective, retaliatory, or otherwise abusive prosecution.^[8] Second, since the president has “exclusive authority” over criminal investigations and prosecutions, any judicial consideration of a claim of abusive prosecution must begin with the president’s own statements and actions, since subordinate prosecutors cannot act contrary to his will; moreover, at least where the president has involved himself in a particular case, he is necessarily part of the “prosecution team” for purposes of the preservation and discovery of evidence — and in any case in which he or others at the White House have received grand jury information, they must be identified to the court pursuant to Federal Rule of Criminal Procedure 6(e)(3)(B). Finally, by acting as chief prosecutor, the president submits himself to the jurisdiction and rules of the relevant court, including with respect to limitations on public statements regarding a pending case. In short, if the president has the powers of a prosecutor, he also has the responsibilities of one.

And when the president deviates from those responsibilities, his actions are not immune from correction. Although Trump has stated that “I have an Article II, where I have the right to do whatever I want as president,”^[9] that is, of course, manifestly untrue. Under Article II, the president is constitutionally required to “take Care that the Laws be faithfully executed,” and by his oath has sworn to “preserve, protect and defend the Constitution of the United States.” The unitary executive theory may unify the Executive but it does not mean the president is unconstrained by the Constitution.^[10] Indeed, as Justice Antonin Scalia explained, the possibility that an Executive may abuse his “exclusive power” to prosecute is checked by the Judiciary’s “exclusive power” to “dismiss malicious prosecutions.”

In sum, just as any other federal prosecutor, the president is a “servant of the law,” and cannot abuse his prosecutorial power or use it as “an instrument of oppression.” Should he do so, as is shown below, courts have the constitutional power and duty to respond, including by holding the president to account as the leader of the prosecution team, quashing subpoenas issued for the purpose of harassment, and — in Justice Scalia’s words — “dismiss[ing] malicious prosecutions.”

I. President Trump Is Acting as Chief Prosecutor of the United States.

In word, policy, and deed, Trump has embraced the role of chief prosecutor of the United States. We start with his public statements. In his first term, Trump asserted “I have [the] absolute right to do what I want to do with the Justice Department.” He also repeatedly called himself the “chief law enforcement officer of the country,” a claim reiterated during his second term.^[11] In addition, during the 2024 campaign, then-candidate Trump stated that if he were re-elected, “based on what [his political opponents have] done, I would have every right to go after them.”

Turning to official policy: following Watergate, successive presidents typically have sought to avoid both appearing to become, and actually becoming, involved in particular criminal cases. To that end, they erected a wall of separation between the White House and the Department of Justice, in particular by establishing a written “contacts policy” limiting who in DOJ can speak with the White House and strictly limiting discussion of particular cases, as opposed to general criminal justice policy issues. During the first Trump administration, commentators raised concerns over apparent violations of the contacts policy by administration officials. And although White House Counsel Don McGahn issued a 2017 memorandum regarding contacts, he addressed it only to White House staff, and its rules did not seem to apply to the president — who, in fact, frequently attempted to intervene in law-enforcement affairs. In the second Trump administration, the White House reportedly circulated internally a memorandum explicitly allowing discussions with the president and vice president about criminal and civil litigation matters.

Finally, as to deeds: Trump has asserted complete authority over the Department of Justice, and has ordered particular investigations and prosecutions. Trump is now literally –– as well as figuratively — the face of the Department of Justice. In one of her first acts, then-Attorney General Pam Bondi issued a memorandum making clear that the Department’s client was the president and that DOJ attorneys were not to exercise independent judgment — a dramatic break from the Department’s prior position.^[12] DOJ followed through on that threat almost immediately, placing three prosecutors on administrative leave following their resistance to a plea deal for then-New York City Mayor Eric Adams. Others have similarly been relieved of their positions or resigned.

The Department also has made clear it is taking its orders directly from the president. For instance, Trump posted on Truth Social last November that he was “asking” Bondi to investigate Democrats related to their alleged involvement with convicted sex offender Jeffrey Epstein. When the president was asked whether it was appropriate to order such investigations, he responded, “I’m the chief law enforcement officer of the country. I’m allowed to do it.” Bondi answered the president’s direction with alacrity, in a post that said “Thank you, Mr. President,” and explained that she had assigned the investigation to U.S. Attorney for the Southern District of New York, Jay Clayton. When Bondi herself was ultimately terminated, apparently for failing to implement the president’s prosecutorial agenda to his satisfaction, Trump nominated Blanche, then-deputy attorney general, to replace her. Blanche, in turn, explicitly refused to pledge that he would act independently from the White House.^[13]

The White House has gone further still, in essence cutting out the Department of Justice entirely as an unnecessary middleman. Earlier this year, Vice President JD Vance announced that the DOJ’s new “Fraud Division” would report to the president and to him. The president has met directly at the White House with U.S. Attorneys from across the country, reportedly “excoriat[ing]” them for failing to pursue cases he cared about more quickly.^[14] And FBI Director Kash Patel apparently was summoned to the White House to plan an investigation into reporting by the New York Times; Patel himself confirmed, “I was at the White House, true, and the fake news will find out why soon.”

But perhaps most telling is the president’s Truth Social post regarding former FBI Director James Comey, New York State Attorney General Letitia James, and U.S. Senator Adam Schiff (D-CA), which bears quoting in full:

Pam: I have reviewed over 30 statements and posts saying that, essentially, “same old story as last time, all talk, no action. Nothing is being done. What about Comey, Adam “Shifty” Schiff, Leticia [sic]??? They’re all guilty as hell, but nothing is going to be done.” Then we almost put in a Democrat supported U.S. Attorney, in Virginia, with a really bad Republican past. A Woke RINO, who was never going to do his job. That’s why two of the worst Dem Senators PUSHED him so hard. He even lied to the media and said he quit, and that we had no case. No, I fired him, and there is a GREAT CASE, and many lawyers, and legal pundits, say so. Lindsey Halligan is a really good lawyer, and likes you, a lot. We can’t delay any longer, it’s killing our reputation and credibility. They impeached me twice, and indicted me (5 times!), OVER NOTHING. JUSTICE MUST BE SERVED, NOW!!! President DJT

Some reports suggest the president meant to send this as a private Direct Message to Bondi, but instead mistakenly posted it publicly. Strikingly, Bondi reportedly was upset not that the president had publicly called for these prosecutions, but that the president’s post suggested she was not acting quickly enough to meet his demands; the president accordingly posted shortly thereafter that she was doing a “GREAT job.” But the message got through: within three weeks, the Department of Justice had indicted both James and Comey, two out of three of the president’s targets, and reportedly was also investigating Schiff.

While the court dismissed on other grounds in the Comey and James cases,^[15] the government’s responses to the defendants’ motions to dismiss on the grounds of vindictive and selective prosecution are telling. Notwithstanding Trump’s Truth Social post, the government argued the defendants failed to establish that the president directed that the prosecutions take place, even while suggesting it would “‘have been surprising’” had the “highest-level officials” not been involved in such a significant case.^[16] Similarly, referring to the president’s Truth Social post, now-Acting Attorney General Blanche explained “[t]hat type of communication from President Trump should make every American happy,” since “[i]t means that there’s an executive, a chief executive, that is making sure every one of his Cabinet members are working as hard as they should.”

Indeed, the Supreme Court suggested in Trump v. United States that this is exactly as it should be under the U.S. Constitution. The Court there held that “[t]he President’s ‘management of the Executive Branch’ requires him to have ‘unrestricted power to remove the most important of his subordinates’—such as the Attorney General—‘in their most important duties.’” And this is true even if the removal is because the subordinates refuse to go along with an abusive prosecution: “[t]he indictment’s allegations that the requested investigations were ‘sham[s]’ or proposed for an improper purpose do not divest the President of exclusive authority over the investigative and prosecutorial functions of the Justice Department and its officials.” So here: the president decided that “[t]here is a GREAT CASE” against Comey and James; he forced the removal of the interim U.S. Attorney who said “we have no case”; and then he put in place a replacement who would prosecute the case “NOW.”

With this Supreme Court imprimatur for his actions, why should Trump hesitate to take full credit for having directly ordered these investigations and prosecutions? Perhaps the government’s half-hearted attempt to suggest that Trump could have, but did not, interfere in these investigations and prosecutions can be interpreted — as has been said of hypocrisy — as the homage that vice pays to virtue. Trump himself has repeatedly used the same technique: suggesting he could have ordered a particular investigation or prosecution but did not do so — notwithstanding all the evidence to the contrary.^[17] This attempt at deflection suggests even the Trump administration recognizes that there is something troubling about the president being seen as personally choosing whom he will have investigated or prosecuted: perhaps not troubling to members of the administration in a moral sense, but in the sense that acknowledging the president’s role as chief prosecutor may trigger consequences they would prefer to avoid. What those consequences are, we discuss in the sections that follow.

II. As Chief Prosecutor, the President May Not Pursue Abusive Prosecutions.

More than 90 years ago, the Supreme Court held that federal prosecutors operate not as “an ordinary party to a controversy,” but instead as a “representative” of “a sovereignty whose obligation to govern impartially is as compelling as its obligation to govern at all.” The “interest therefore” of the U.S. government “in a criminal prosecution is not that it shall win a case, but that justice shall be done.” The Court stated that the federal prosecutor is thus “in a peculiar and very definite sense the servant of the law, the twofold aim of which is that guilt shall not escape or innocence suffer.”

The Constitution’s Due Process Clause compels this conclusion. As the Supreme Court has repeatedly held, the Due Process Clause prevents government officials “‘from abusing [their] power, or employing it as an instrument of oppression.’” In particular, it forbids retaliatory or vindictive prosecutions based on an individual’s exercise of his rights: “[t]o punish a person because he has done what the law plainly allows him to do is a due process violation of the most basic sort . . . and for an agent of the State to pursue a course of action whose objective is to penalize a person’s reliance on his legal rights is ‘patently unconstitutional.’” Likewise, “while an individual certainly may be penalized for violating the law, he just as certainly may not be punished for exercising a protected statutory or constitutional right.”

Moreover, abusive prosecutions may violate not just the Due Process Clause, but other provisions of the Constitution as well. Thus, “the First Amendment prohibits government officials from subjecting an individual to retaliatory actions, including criminal prosecutions, for speaking out.” And the Equal Protection Clause prohibits the government from selectively deciding “whether to prosecute . . . based on ‘a standard such as race, religion, or other arbitrary classification’” or from “intentionally treat[ing]” the defendant “differently from others similarly situated” where “there is no rational basis for the difference in treatment.”

But the Due Process Clause protects not just against selective federal prosecutions or prosecutions brought to punish an individual for exercising his or her rights. More generally, it forbids any prosecution brought “as an instrument of oppression,” including a prosecution brought for the purpose of gratifying the animus — or whim — of the prosecutor. As the Supreme Court has held, it is “a fundamental premise of our society that the state wield its formidable criminal enforcement powers in a rigorously disinterested fashion.”^[18]

Several federal district courts have found that the second Trump administration has in fact abused its prosecutorial power by unlawfully undertaking criminal investigations or prosecutions to harass, punish, or pressure the administration’s opponents. In In re Grand Jury Subpoenas Nos.__ & __, Chief Judge James Boasberg of the United States District Court for the District of Columbia quashed two subpoenas issued to the Federal Reserve, holding that “[i]n light of all the evidence” –– which included social media statements by Trump attacking Federal Reserve Chair Jerome Powell — “the only reasonable inference is that the Government targeted Powell ‘out of malice or an intent to harass’ and has launched a ‘fishing expedition[]’ to either find something to pin on him or to pressure him to fold” because of his management of the Federal Reserve.^[19] In In re Grand Jury Subpoenas, Chief Judge Patrick Schiltz of the United States District Court for the District of Minnesota quashed subpoenas issued to the offices of Minnesota state and local officials, finding –– based on “overwhelming” evidence that included Trump’s social media statements –– that the subpoenas were part of a criminal investigation undertaken to “harass political opponents or to coerce them into taking official action,” which is a “blatantly unlawful and unethical use [of] the grand-jury process.”^[20] And in United States v. Abrego Garcia, Judge Waverly Crenshaw of the United States District Court for the Middle District of Tennessee dismissed a criminal prosecution, finding the government failed to rebut the presumption that the prosecution was vindictively brought to punish the defendant for asserting his rights; specifically, the court found “persuasive, credible, objective evidence” contrary to the government’s claim that the prosecutor decided to indict “free of influence, pressure, or direction from [then-Deputy Attorney General] Blanche, Main Justice, the White House, or DHS.” The government has appealed this holding.

In sum, even if under Trump v. United States the president can punish subordinates for not pursuing prosecutions for abusive purposes, that does not mean he can pursue such prosecutions without violating the Constitution.

III. In Considering Abusive Prosecution Claims, Courts Should Treat the President’s Actions and Statements as Those of the Chief Prosecutor of the Case.

Whether in any particular case the president has directed that a prosecution proceed for an abusive purpose is a question of fact for a court to decide. But given the Supreme Court’s ruling in Trump v. United States that the president has “exclusive authority over the investigative and prosecutorial functions of the Justice Department and its officials” –– and the ample evidence noted above that Trump has avidly embraced and exercised that authority — a court reviewing a defendant’s abusive prosecution claim must begin its analysis by considering any public accusations against the defendant by the president and must treat such statements as providing direct evidence — or at a minimum, as creating a rebuttable presumption — that the president himself made the decision to prosecute.^[21]

In abusive prosecution cases, courts frequently consider whether the prosecutor reached an independent decision to prosecute.^[22] But presidential involvement makes that analysis inapposite, since under the unitary executive theory, once the president has determined a prosecution should be pursued, a prosecutor may not independently decide not to go forward. As the Supreme Court explained last term in Trump v. Slaughter:

To vest “the whole executive power” in just one person was not to suggest that he could execute the laws alone and unaided. The Federalist No. 47, at 303 (J. Madison). But it was to establish a hierarchy—a “Chief Magistrate” with whom the buck stops, and below him various “assistants or deputies” who “derive their offices from his appointment” and remain “subject to his superintendence.”

Thus, the Court explained, the president’s “so-called assistants” cannot “exercise his power against his wishes,” since this would not allow “the Constitution [to] live up to James Iredell’s boast that ‘the President’ would be ‘personally responsible for everything.’” This applies with particular force in the context of litigation since “the ‘discretionary power to seek judicial relief’ lies at the very core of executive authority,” and thus it is “up to the Executive to ‘decide how to prioritize and how aggressively to pursue legal actions against defendants who violate the law.’”

In other words, if the president directs a prosecution to proceed, it is irrelevant whether the line prosecutor who appears in court claims to have subsequently made an independent assessment of the case’s validity. Under the unitary executive theory adopted in Trump v. United States — and as reinforced by the Bondi memorandum requiring DOJ prosecutors to pursue cases regardless of their personal judgment — a federal prosecutor has no independent authority to assess whether a case is proper. As a result, if the president’s decision to prosecute was based on unlawful considerations, it cannot be “cured” by the actions or statements of subordinate prosecutors, since they serve only to effectuate his will.

To be sure, federal law makes the “Department of Justice, under the direction of the Attorney General,” responsible for federal litigation^[23] and further gives U.S. Attorneys the duty to prosecute federal crimes within their districts. Likewise, the Federal Rules of Criminal Procedure define who is an “attorney for the government” for the purposes of “criminal proceedings” in the federal courts.^[24] But although the president may not be the “attorney for the government” who appears in court, the Supreme Court has held that his power “to decide which crimes to investigate and prosecute” is “exclusive . . . and absolute” and that he “may discuss potential investigations and prosecutions with his Attorney General and other Justice Department officials to carry out his constitutional duty to ‘take Care that the Laws be faithfully executed’” (emphasis added).^[25] Indeed, Professor Prakash argues that “[t]aken together, the Executive Power and Faithful Execution Clauses suggest that the president may direct official prosecutors and, in some cases, has a duty to do so. . . . The president’s faithful execution duty may require him to use his executive power to direct the official prosecutors.”^[26]

Thus, when evidence suggests the president has involved himself in a particular case, a court considering an abusive prosecution claim must treat the president’s statements and actions as those of the chief prosecutor in that case. This means, in the first instance, that the court must consider whether the president’s statements or actions in and of themselves constitute sufficient proof of impermissible purposes.^[27] Alternatively, the court must consider whether those statements or actions demonstrate “a realistic likelihood of vindictiveness” or other forbidden abusive purpose, which the government then must rebut.^[28]

Evidence that the president has chosen to intervene in a particular prosecution is not necessarily limited to his own public statements or actions. Such involvement also may be signaled by his subordinates’ words or deeds or by structural changes the administration has made to bring prosecutions directly under White House control.^[29] Furthermore, such involvement may be evinced by the sharing of grand jury information with the White House. Under Federal Rule of Criminal Procedure 6(e), a matter occurring before a grand jury may be shared with “any government personnel . . . that an attorney for the government considers necessary to assist in performing that attorney’s duty to enforce federal criminal law.” Thus, when acting as chief prosecutor, the president may receive such information. But anyone to whom such information has been disclosed “may use that information only to assist an attorney for the government in performing that attorney’s duty to enforce federal criminal law.” Further, “[a]n attorney for the government must promptly provide the court that impaneled the grand jury with the names of persons to whom a disclosure has been made” and must certify that the recipients have been advised of their obligation to preserve the secrecy of this information. In sum, if any grand jury information has been shared with the president or any other White House personnel, the Department of Justice must inform the court.^[30] Such disclosure necessarily establishes the president’s involvement as prosecutor in a particular case — since sharing such evidence without a reason that would “assist an attorney for the government in performing that attorney’s duty to enforce federal criminal law” –– or to protect national security –– would violate grand jury secrecy.^[31]

Once the president’s involvement in a particular investigation or prosecution is established, other obligations follow, including with respect to the preservation and discovery of evidence. As the DOJ Manual explains, “the prosecution team” for purposes of discovery includes “federal, state, and local law enforcement officers and other government officials participating in the investigation and prosecution of the criminal case” (emphasis added). Where the president is involved in a particular case, he is by definition the head of the “prosecution team.” The line prosecutor accordingly has the “obligation[,] . . . in preparing for trial, to seek all exculpatory and impeachment information from all the members of the prosecution team,” including the president or any other White House officials who participated in the investigation or prosecution.

In addition, the Justice Manual provides that “[a]ll prosecution team members should be aware of the government’s obligations regarding the preservation and disclosure of electronic communications,” including that “case-related e-communications may potentially be disclosed to the defense.” Thus, the Manual continues, “all team members should understand . . . the obligation to preserve and produce such communications when appropriate.” The Justice Manual further provides that “[p]rosecution team members should not use personally owned electronic communication devices, personal email accounts, social networking sites, or similar accounts to transmit case-related information to witnesses or other team members.” It also makes clear that this “guidance applies at all phases of a criminal case including investigation, trial preparation, trial, and post-trial” (emphasis added). As one commentator has noted, if, for instance, Trump is considered part of the prosecution team for the Comey prosecution, “[a]t a minimum,” DOJ may need to take steps to preserve his electronic communications, including from personal devices.^[32]

Following this reasoning, the United States District Court for the Southern District of Texas recently concluded that discovery in a criminal case could extend to possible communications by Trump, given his control over the Department of Justice. In United States v. Yuan, the defendant — charged with conspiring to export-controlled computer chips to China — sought discovery under Federal Rule of Criminal Procedure 16 and Brady of published or unpublished statements by Trump relating to the enforcement of the statutes involved in his case, including any relevant communications with the president of China. The court directed discovery, holding:

The Presidency and the Department of Justice . . . constitute one branch of government . . . . The President is not exempt from the law (FRCP, Rule 16), even though it may be argued that he cannot be prosecuted for violations.

In its Motion to Reconsider, the Department of Justice — including its National Security Division — argued the information sought was shielded by various privileges, and that in any event the president could not be considered part of the prosecution team in that particular case, because he had not been involved in it. The Department stated that while “[t]he President is the head of the executive branch of which the Department of Justice is part,” he was not part of the “investigative and prosecutorial personnel” that brought the case: “[t]he President was not involved in the investigation or prosecution of the Defendant and is therefore not a part of the prosecution team” (emphasis added). This necessarily concedes that the president would be part of the prosecution team where he was involved in an investigation or prosecution.^[33]

IV. When Acting as Chief Prosecutor, the President Submits Himself to the Jurisdiction of the Court.

When federal prosecutors bring a case, they submit themselves to the jurisdiction of the court hearing the case. The prosecutor accordingly becomes subject both to the general Federal Rules of Criminal Procedure and the local rules of that court. The same principles must apply also when the president acts as chief prosecutor in a particular case: as leader of the prosecution team, the president must be subject to the court’s jurisdiction and rules to the same extent as the rest of the team.

This does not present a separation of powers issue. After all, as noted above, under Trump v. United States, and in accord with the unitary executive theory, the line prosecutor who appears in court simply stands in for the president as his “assistant.” The trial prosecutor thus has no independent will or power; both derive entirely from the president. The same applies even to the U.S. Attorney and the attorney general — they too derive all their authority from the president. Yet no one would assert that the unitary executive theory implies a court could not assert jurisdiction over the line prosecutor in a particular case, or — as appropriate — over those in the prosecutor’s chain of command in the Department of Justice. It follows that no separation of powers issue arises from the extension of jurisdiction to the president when he acts as chief prosecutor in a particular case.^[34] To conclude otherwise would suggest that the unitary executive theory goes far beyond unifying the executive, instead destroying the separation of powers by giving the president and his subordinates prosecutorial powers the Judiciary cannot constrain.

Justice Scalia’s dissent in Morrison v. Olson addressed this point, noting that the unitary executive theory, while centralizing all executive power in the hands of the president, does not eliminate judicial checks on the exercise of this power. Scalia took the position that “the President’s constitutionally assigned duties include complete control over investigation and prosecution of violations of the law” — the very position the Supreme Court subsequently adopted in Trump v. United States. He then asked rhetorically: “Is it unthinkable that the President should have such exclusive power, even when alleged crimes by him or his close associates are at issue?” Scalia answered his own question by suggesting that “[a] system of separate and coordinate powers necessarily involves an acceptance of exclusive power that can theoretically be abused.” But he went on to note that the check “against any branch’s abuse of its exclusive powers” is in the first instance “retaliation by one of the other branch’s use of its exclusive powers”: in particular, “the courts can dismiss malicious prosecutions” by the president.^[35]

Thus, once the president acts as chief prosecutor by directing that a prosecution take place, he cannot complain that he must answer to a court like any other prosecutor and obey that court’s rules. This includes, where local rules provide, restrictions on prejudicial pretrial statements. In United States v. Abrego Garcia, for instance, the District Court found that government officials — including Bondi and then-Department of Homeland Security Secretary Kristi Noem — publicly “made allegations regarding [defendant’s] ‘character or reputation,’” expressed views on his “‘guilt or innocence,’” and “offer[ed] an opinion ‘as to the evidence in the case’” — all prohibited by Local Criminal Rule 2.01(a)(2)(B).^[36] Accordingly, the court granted the defendant’s motion “to apply [the rule] to ‘order that all DOJ and DHS officials involved in this case, and all officials in this supervisory chain, including Attorney General Pamela Bondi and Secretary of Homeland Security Kristi Noem, refrain from making extrajudicial comments that pose a substantial likelihood of materially prejudic[ing] this proceeding.’”^[37]

Other federal courts have likewise extended restrictions on prejudicial statements to government officials beyond the line prosecutors conducting the trial. In United States v. Mangione, the United States District Court for the Southern District of New York found that “multiple employees at the Department of Justice may have violated Local Criminal Rule 23.1,” which bars statements by a lawyer or law firm — or non-lawyer personnel employed by the lawyer’s office or subject to a lawyer’s supervision — where there is a “substantial likelihood [those statements] will interfere with a fair trial or otherwise prejudice the administration of justice.”^[38] The court further instructed the government “to advise the Deputy Attorney General, for dissemination within the Department as appropriate, that future violations may result in sanctions, which could include personal financial penalties, contempt of court findings, or relief specific to the prosecution.”^[39]

Although a president might seek to distinguish Mangione on the grounds that the local rule at issue there applies only to lawyers and “non-lawyer personnel” employed or supervised by a lawyer, the broader point remains that courts may limit prejudicial speech by those associated with the prosecution team.^[40] Over two decades ago, the United States District Court for the Eastern District of Michigan made the same point in United States v. Koubriti when it “formally and publicly admonished” then-Attorney General John Ashcroft for comments that violated an earlier order prohibiting statements with a substantial likelihood of prejudicing the defendants in a terrorism trial.^[41] The court “fully recognize[d]” that “the Attorney General occupies two roles of equal importance, one as the Nation’s chief prosecutor, and one as the head of an Executive department with responsibilities to keep the public informed on policy matters.”^[42] It also recognized that “the Attorney General undoubtedly faces a difficult and challenging task in harmonizing these two roles,” yet it stated:

[I]t is unquestionably the duty of this Court to ensure that the defendants who appear before it are accorded their fair trial rights under the Constitution, to enter orders designed to protect these rights, and to see that its orders are obeyed.^[43]

The same reasoning perforce applies when the president chooses to displace the attorney general as “the Nation’s chief prosecutor” in a particular case: he, too must obey the orders of the court. Whether a court can enforce its orders against the president by means of contempt proceedings presents separation of powers issues beyond this article’s scope.^[44] But at least with respect to criminal contempt, the separation of powers issues do not arise because of the president’s involvement per se. Rather, there is an open question whether a district court may pursue a criminal contempt prosecution against any federal official without the participation of the Department of Justice. As a matter of logic, assuming district courts can pursue criminal contempt against prosecutors who disobey their orders, they should also be able to do so when the president acts as chief prosecutor. But here Trump v. United States may triumph over logic, affording the president immunity from criminal prosecution even when he has submitted himself to the jurisdiction of the court by acting as chief prosecutor in a particular case.^[45]

Regardless whether a court may hold a president in criminal contempt, other sanctions remain available: quashing criminal subpoenas or dismissing prosecutions clearly pursued, in the words of the Supreme Court, as an “instrument of oppression.” As noted above, several courts have done precisely this.^[46] Such judicial action, moreover, fully accords with the unitary executive theory, which assures us, after all, that the president cannot hide from the consequences of his actions, or divert responsibility for unlawful measures onto his assistants. To the contrary, as the Supreme Court recently stated, “[w]ith just one President in charge, [the Framers] reasoned, there would be no doubt ‘on whom the blame or the punishment of a pernicious measure . . . ought really to fall.” Nor is it just voters who can sit in judgment in this regard: federal courts can –– indeed must –– determine whether they are being asked to further prosecutions unconstitutionally advanced by the president. As the United States District Court for the Southern District of Florida recently explained, in connection with the purported settlement of Trump’s civil lawsuit against the IRS:

The President may be the functional “dominus litus” [sic] of the Executive Branch, but as a party to a civil suit, he, as well as all the parties and lawyers before a court, are bound by the rules. Ensuring that our courts are used only for the express purpose created by the Constitution is the obligation of every judge and an obligation that this Court must discharge in light of the matter before it.^[47]

This reasoning applies with equal, if not greater, force in the criminal context.

CONCLUSION

The Supreme Court has established the president as chief prosecutor of the United States, and Trump has embraced the role. But this does not confer power without constraint. Acting in this role, the president must continue to “take care that the Laws be Faithfully Executed” and must safeguard the rights of individuals to “due process of law” when their life, liberty, or property is subject to potential deprivation. The unitary executive theory may lead to the conclusion that the president is the chief prosecutor, but it cannot be read to grant him powers beyond those the Constitution grants to execute that function. To the contrary, when the president purports to act as chief prosecutor, he must abide by the same constitutional, legal, and ethical constraints that guide all federal prosecutors. In sum, when a president becomes a prosecutor — when he decides who will be investigated and prosecuted — he necessarily becomes “the servant of the law” and “the representative . . . of a sovereignty whose obligation to govern impartially is as compelling as its obligation to govern at all.”

The authors are grateful for the advice of Professors of Law Kate Stith (Yale), Harold Hongju Koh (Yale) and Steven Arrigg Koh (Boston University).

– – – – – – – – – – – – – – –

1. See discussion in Part I infra. ↑
2. See discussion in Part I infra. ↑
3. See Trump v. United States, 603 U.S. 593, 609, 620–21 (2024) (ruling that the President has “conclusive and preclusive” power “to decide which crimes to investigate and prosecute,” and has “exclusive authority over the investigative and prosecutorial functions of the Justice Department and its officials”); see also Saikrishna Bangalore Prakash, The Chief Prosecutor, 73 George Washington L. Rev., 521, 521 (2005), https://digital.sandiego.edu/lwps_public/art30 (noting academic debate over whether prosecution is an exclusively executive function, but arguing that “the Constitution’s original understanding, constitutional text, structure, and history establish that the President is the constitutional prosecutor of all federal offenses . . . .”). ↑
4. See also Rob Gillies, Canadian Ethics Commissioner Finds Justin Trudeau Violated Ethics, PBS News (Aug. 14, 2019), https://www.pbs.org/newshour/world/canadian-ethics-commissioner-finds-justin-trudeau-violated-ethics (describing Canadian Ethics Commissioner’s finding that then-Prime Minister Justin Trudeau’s attempts to influence former attorney general to halt prosecution of a company violated prosecutorial independence). ↑
5. See Prakash, supra note 3 at 539 (arguing that the President not only may act as chief prosecutor, but that “[t]he president’s faithful execution duty may require him to use his executive power to direct the official prosecutors”) (emphasis added). But see infra nn.6-7. ↑
6. See also Caleb Nelson, Against the Dictatorial Presidency 65-68 (Sep. 8, 2026) (unpublished manuscript), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7365761 (arguing that the unitary executive theory “has encouraged President Trump’s assertions of control over specific prosecutorial decisions at the Department of Justice” in a way that “violates norms” while “the Supreme Court is making it difficult for Congress to codify the norms in a way that could restrain a President” and listing examples); Michael R. Dreeben, Robert Jackson’s The Federal Prosecutor Revisited, 139 Harv. L. Rev. F. 173, 174 (2026) (“The Supreme Court’s modern view of the unitary executive provides a roadmap for upending the Department’s traditions of independence and impartiality; it invites the Executive to deploy the Department’s power for partisan ends.”). ↑
7. For the contrary view, see Bruce A. Green & Rebecca Roiphe, Can the President Control the Department of Justice, 70 Ala. L. Rev. 1, 75 (2018) (“The Constitution does not determine whether ultimate authority rests with the President or the prosecutors, leaving Congress to decide.”). Compare Morrison v. Olson, 487 U.S. 654, 669, 695 (1988) (Rehnquist, C.J., for the Court) (upholding Independent Counsel law notwithstanding that it is “undeniable that the Act reduces the amount of control or supervision that the Attorney General and, through him, the President exercises over the investigation and prosecution of a certain class of alleged criminal activity”) with id. at 726 (Scalia, J., dissenting) (describing “a prosecutor” as “the virtual embodiment of the power to ‘take care that the laws be faithfully executed’”). ↑
8. This article uses the term “abusive prosecutions” to encompass cases unconstitutionally brought as “instrument[s] of oppression.” See discussion in Part II infra; see also Morrison v. Olson, 487 U.S. 654, 711 (1988) (Scalia, J., dissenting) (arguing the executive has “exclusive power” to prosecute, but courts may dismiss “malicious prosecutions”); Jeffrey Bellin, Faithless Prosecution, 95 Geo. Wash. L. Rev. (forthcoming 2027) (calling the vindictive and selective prosecution doctrines “undertheorized” and “poorly suited to address the new problem of a politicized DOJ” and urging courts “to recognize a third constitutional limit on charging based in the Take Care Clause” and thus develop “a doctrine barring ‘faithless prosecutions’”) (manuscript at 5, 48). ↑
9. See also Amanda Terkel & Lawrence Hurley, Trump, Asked if He Has to ‘Uphold the Constitution,’ Says, ‘I Don’t Know,’ NBC News (May 4, 2025), https://www.nbcnews.com/politics/trump-administration/trump-asked-uphold-constitution-says-dont-know-rcna204580 (quoting President Trump as responding “I don’t know” when asked if he has to uphold the Constitution). ↑
10. Although in the just-completed Supreme Court term, Trump v. Slaughter, 609 U.S. ___ (2026) moved constitutional law closer to the model of a unitary executive, Trump v. Cook, 609 U.S. ___ (2026) showed that this unity is not complete. But regardless of the reach of the theory, and no matter how much power it confers on the President, it cannot justify the President’s use of that power in an unconstitutional manner. ↑
11. See also Farrah Tomazin, Trump Wrongly Claims He’s Crime Chief as He Blasts ‘Unpatriotic’ Bolton, Yahoo News (Aug. 22, 2025), https://www.yahoo.com/news/articles/trump-wrongly-claims-crime-chief-163741553.html (similar); Nicholas Riccardi, Trump Ran on a Promise of Revenge. He’s Making Good on It, NBC News (Aug. 23, 2025), https://www.nbcwashington.com/news/national-international/trump-ran-promise-revenge-making-good/3979201/(“I could know about it. I could be the one starting it,” the President told reporters. “I’m actually the chief law enforcement officer.”) ↑
12. The Department of Justice’s historical credo is carved into the rotunda of the Attorney General’s Office: “The United States Wins Its Point Whenever Justice Is Done Its Citizens in the Courts.” Josh Geltzer, Law Day and the Need for Virtue in Government Lawyering, Just Secur. (May 1, 2020), https://www.justsecurity.org/69971/law-day-and-the-need-for-virtue-in-government-lawyering/. ↑
13. See also Ali Rogin & Jonah Anderson, Senators Question Blanche on Trump Loyalty, Jan. 6 Cases and Epstein Files, PBS News (July 15, 2026), https://www.pbs.org/newshour/show/senators-question-blanche-on-trump-loyalty-jan-6-cases-and-epstein-files (discussing questions about Blanche’s personal relationship with President Trump). ↑
14. In re Grand Jury Subpoenas Nos. __ & __, 823 F. Supp. 3d 1, 6 (D.D.C. 2026). ↑
15. The indictments in the Eastern District of Virginia against Comey and James were dismissed because Lindsey Halligan, who obtained the indictments, was unlawfully appointed as Interim U.S. Attorney. United States v. James, 810 F. Supp. 3d 752, 754 (E.D. Va. 2025); United States v. Comey, 810 F.Supp.3d 768, 771 (E.D. Va. 2025). Reportedly, another grand jury in the Eastern District of Virginia twice rejected the government’s subsequent attempt to reindict James. Hannah Rabinowitz et al., Justice Department Fails to Reindict Letitia James for a Second Time, CNN (Dec. 11, 2025), https://www.cnn.com/2025/12/11/politics/justice-department-fails-reindict-letitia-james-second-time. Comey was subsequently indicted on different charges in the Eastern District of North Carolina. Alanna Durkin Richer & Eric Tucker, Ex-FBI Director Comey Indicted Again, in a Probe Over an Online Post Officials Call a Trump Threat, Associated Press (Apr. 28, 2026), https://apnews.com/article/comey-indicted-seashell-photo-86-47-a7fdd67891a7f74bc6fd8ce4d3d4170a. ↑
16. Compare United States’ Response in Opposition to Defendant’s Motion to Dismiss Indictment Based on Vindictive and Selective Prosecution at 16, United States v. Comey, No. 1:25-CR-272 (E.D. Va. Nov. 3, 2025) (quoting President Trump’s response when asked whether an indictment of Comey was imminent: “I can’t tell you what’s going to happen because I don’t know yet. Very professional people, headed up by the Attorney General, [Deputy Attorney General] Todd Blanche and [U.S. Attorney] Lindsey Halligan . . . they’re going to make a determination. I’m not making that determination. I think I’d be allowed to get involved if I want, but I don’t really choose to do so.”) with id. at 40 (“Indeed, given the nature of the allegations lodged against a former Executive Branch agency head, ‘it would have been surprising had the highest-level officials not been involved in the decision to prosecute’”) (quoting United States v. Lindh, 212 F. Supp. 2d 541, 568 (E.D. Va. 2002)). ↑
17. See also Malcolm Ferguson, Trump Makes Clear No One Can Rule Against Him as He Threatens Enemies, The New Republic (Oct. 15, 2025), https://newrepublic.com/post/201828/trump-chief-law-enforcement-officer-threatens-enemies (“And I’m allowed to find out [about investigations], I’m allowed, in theory, the ‘chief law enforcement officer’”). ↑
18. Similarly, pointing to the Take Care Clause, Professor Bellin has called for courts to recognize, and void, “faithless prosecutions” by the President or those he directs. See supra note 8. ↑
19. In re Grand Jury Subpoenas Nos. __ & __, 823 F.Supp.3d 1, 14 (D.D.C. 2026) (quoting United States v. R. Enters., Inc., 498 U.S. 292, 299 (1991)). ↑
20. In re Grand Jury Subpoenas, 835 F.Supp.3d 947, 955 (D. Minn. 2026); cf. Fulton Cnty. Bd. of Registration and Elections v. United States, 836 F.Supp.3d 1305, 1309 n.2, 1315 (N.D. Ga. 2026) (holding that “in pursuing the Subpoena [for records relating to election workers] the DOJ is engaged in an ‘arbitrary fishing expedition,’ such that the Subpoena is unreasonable and must be quashed” and noting that “no evidence has been presented to the Court that the actual Grand Jury in the Northern District of Georgia seeks this information, as opposed to the out-of-district prosecutors who the DOJ has appointed to lead this inquiry who have served this Subpoena in the name of the Grand Jury.” (quoting Trump v. Vance, 591 U.S. 786, 805 (2020))). ↑
21. Courts may take judicial notice of the public statements of the President and other officials. See, e.g., cases cited at notes 19-20, supra; Trump v. IRS, 836 F.Supp.3d 1225, 1233 n.3 (S.D. Fla. 2026) (“The Court sua sponte takes judicial notice of articles reporting on relevant events attendant to this case, dockets of other cases, and statements the Parties made. . . . Moreover, the Department of Justice has directed courts to look to unsworn public statements as binding party admissions.”). ↑
22. See, e.g., United States v. Koh, 199 F.3d 632, 640 (2d Cir. 1999) (explaining that a defendant can make out a vindictive prosecution claim by showing that the prosecutor “was prevailed upon to bring the charges by another with animus such that the prosecutor could be considered a ‘stalking horse.’”) (quotation marks and citation omitted); United States v. Schoolcraft, 879 F.2d 64, 68 (3d Cir.) (rejecting claim of vindictive prosecution because defendant failed to allege that anyone at the charging office bore him any impermissible ill-will and because he had not demonstrated that the decision to prosecute him “was not based on the ‘usual determinative factors’”); United States v. Monsoor, 77 F.3d 1031, 1035 (7th Cir. 1996) (holding that in order to make out a vindictive prosecution claim, “[a] defendant must show that the ill will, whoever its bearer, actually motivated his prosecution” and that the referring entity “prevailed upon the prosecutor in making the decision to seek an indictment”). ↑
23. See also 28 U.S.C. § 519 (2024) (directing the Attorney General to “supervise all litigation” and “direct all United States attorneys . . . in the discharge of their respective duties”). ↑
24. Fed. R. Crim. P. 1(a)(1), 1(b)(1) (defining “attorney for the government” to include “(A) the Attorney General or an authorized assistant; (B) a United States attorney or an authorized assistant; (C) when applicable to cases arising under Guam law, the Guam Attorney General or other person whom Guam law authorizes to act in the matter; and (D) any other attorney authorized by law to conduct proceedings under these rules as a prosecutor”). ↑
25. As Chief Justice Roberts further explains:And the Attorney General, as head of the Justice Department, acts as the President’s “chief law enforcement officer” who “provides vital assistance to [the President] in the performance of [the President’s] constitutional duty to ‘preserve, protect, and defend the Constitution.’”Trump v. United States, 603 U.S. 593, 620 (2024) (quoting Mitchell v. Forsyth, 472 U.S. 511, 520 (1985) (quoting U.S. Const. Art. II, § 1, cl. 8)) (emphasis added). ↑
26. Prakash, supra note 3, at 1721; see also id. at 1722 n. 125 (explaining that nothing “in the Constitution requires Congress to provide the president prosecutorial support,” but that “once Congress creates official prosecutors, they are under the president’s direction because they help carry into execution the president’s powers.”); Saikrishna Bangalore Prakash, Too Unitary, 135 Yale L.J.F. 533, 552-53 (2026) (“The President, and not the Attorney General, is the chief law-enforcement officer of the United States. . . . [T]he Constitution authorizes the President to superintend and control the legal interpretations of his mere instruments. But . . . . he lacks constitutional power to delegate his constitutional powers.”). ↑
27. No prosecutor — including the President — can exercise “prosecutorial discretion” to engage in an abusive prosecution. As the Supreme Court has made clear, “‘the conscious exercise of some selectivity in enforcement is not in itself a federal constitutional violation’ so long as ‘the selection was [not] deliberately based upon an unjustifiable standard such as race, religion, or other arbitrary classification.’” Bordenkircher v. Hayes, 434 U.S. 357, 364 (1978) (quoting Oyler v. Boles, 368 U.S. 448, 456 (1962)). See U.S. Dep’t of Just., Just. Manual § 9-27.220 (2023) (setting out the factors that a federal prosecutor must consider before bringing charges); id. at § 9-27.260 (describing “[i]mpermissible [c]onsiderations” when deciding whether to bring charges, including: “[t]he person’s race, religion, gender, ethnicity, national origin, sexual orientation, or political association, activities, or beliefs”; “[t]he attorney’s own personal feelings concerning the person, the person’s associates, or the victim”; or “[t]he possible effect of the decision on the attorney’s own professional or personal circumstances”). ↑
28. See United States v. Goodwin, 457 U.S. 368, 373 (1982) (“Motives are complex and difficult to prove. As a result, in certain cases in which action detrimental to the defendant has been taken after the exercise of a legal right, the Court has found it necessary to ‘presume’ an improper vindictive motive. Given the severity of such a presumption, however . . . the Court has done so only in cases in which a reasonable likelihood of vindictiveness exists.”). ↑
29. See, e.g., Graig Graziosi, Kash Patel Instructed by the White House to Oversee Investigation into NYT Journalists, Indep. (July 11, 2026), https://www.the-independent.com/news/world/americas/us-politics/trump-kash-patel-nyt-investigation-b3013312.html (quoting FBI Director Kash Patel as confirming, regarding the investigation of journalists, “I was at the White House, true, and the fake news will find out why soon”); Alan Feuer, Vance Announces New Justice Dept. Fraud Post to Be ‘Run Out of the White House,’ N.Y. Times (Jan. 8, 2026), https://www.nytimes.com/2026/01/08/us/jd-vance-justice-dept-fraud-post.html (reporting on plans for a new “high-ranking position” at the DOJ “with broad authority to investigate fraud across the country” set to “answer directly” to the President and Vice President and “run out of the White House”). ↑
30. Cf. The Southern Poverty Law Center’s Motion to Show Cause Regarding the Superseding Indictment at 5-10, 13, United States v. S. Poverty L. Ctr., No. 2:26-cr-0139 (M.D. Ala. June 3, 2026) (alleging that DOJ’s release to the media of an unsigned copy of a superseding indictment violated Fed. R. Crim. P. 6(e)’s secrecy provisions and the Justice Manual’s confidentiality policy and requesting a show cause hearing to explore the possibility for sanctions). ↑
31. Fed. R. Crim. P. 6(e)(3)(B); see also Disclosure of Grand Jury Matters to the President and Other Officials, 17 Op. O.L.C. 59, 59 (1993) (concluding that grand jury information may be shared with the President and National Security Council officials where necessary to assist in the enforcement of federal criminal law, or to protect national security, but emphasizing requirement to provide notice of such disclosures to the court). ↑
32. See also Alan Feuer, Ex-C.I.A. Chief Asks for Order Forcing Trump Administration to Preserve Records, N.Y. Times (July 1, 2026), https://www.nytimes.com/2026/07/01/us/politics/john-brennan-investigations-records.html (describing request filed by former CIA Director John Brennan’s legal team to require prosecutors to preserve records for future discovery “in anticipation of bringing a motion for vindictive and selective prosecution if charges against Mr. Brennan are ultimately filed”). ↑
33. It is not the law, as the government suggested in its Motion to Reconsider in Yuan, that to be part of the prosecution team one must have been “‘actively involved in trial preparation, [the] production of evidence, [or the] examination of witnesses’” or have “‘work[ed] hand in hand with’ law enforcement in investigating and ‘presenting [Defendant’s] case before the courts.’” Motion to Reconsider at 1, United States v. Yuan, No. 4:25-cr-00687 (S.D. Tex. June 9, 2026) (quoting Fierro v. Johnson, 197 F.3d 147, 155-56 (5th Cir. 1999)). Rather, as the Justice Manual makes clear, the team includes all “federal, state, and local law enforcement officers and other government officials participating in the investigation and prosecution of the criminal case.” U.S. Dep’t of Just., Just. Manual § 9-5.001 (2020) (citing Kyles v. Whitley, 514 U.S. 419, 437 (1995)). While the court did grant the motion to reconsider in Yuan it did so on the ground that “the government[’]s response …that identifies BIS documents and serves them upon the defendant . . . constitutes the relevant materials available to the prosecution team.” ↑
34. Regardless of whether the unitary executive theory posits that the President must act as Chief Prosecutor, see Prakash supra note 3, at 1721-22, or instead may choose to do so in particular matters while otherwise delegating responsibility to the Attorney General, in either case the Executive voluntarily submits itself to the Judiciary by choosing to go forward with a prosecution. ↑
35. See also Prakash, supra note 26, at 545 (“The President must honor judicial judgments, a duty immanent in the separation of judicial power from executive power.” (first citing William Baude, The Judgment Power, 96 Geo. L.J. 1807 (2008); then citing Gary Lawson & Christopher D. Moore, The Executive Power of Constitutional Interpretation, 81 Iowa L . Rev. 1267, 1303-13 (1996); and then citing Edward A. Hartnett, A Matter of Judgment, Not a Matter of Opinion, 74 N.Y.U. L. Rev. 123, 147-48, 152-54 (1999))). ↑
36. United States v. Abrego Garcia, 350 F.R.D. 557, 562-63 (M.D. Tenn. 2025). ↑
37. Id. at 561. ↑
38. United States v. Mangione, 25-CR-00176, 2025 WL 2715733, at *1 (S.D.N.Y. Sep. 24, 2025). ↑
39. Id. ↑
40. The Local Rules for the United States District Courts for the Southern and Eastern Districts of New York — like those in many other federal district courts — state that “[t]he court, on motion of either party or on its own motion, may issue a special order governing matters such as extrajudicial statements by parties and witnesses likely to interfere with the rights of the accused to a fair trial by an impartial jury.” S.D.N.Y. & E.D.N.Y. Loc. Crim. R. 23.1(h). An extrajudicial statement by the President acting as Chief Prosecutor would be a statement by a “party” within the meaning of this provision. ↑
41. United States v. Koubriti, 305 F. Supp. 2d 723, 765 (E.D. Mich. 2003). ↑
42. Id. at 760. ↑
43. Id. ↑
44. See Nicholas R. Parrillo, The Endgame of Administrative Law: Governmental Disobedience and the Judicial Contempt Power, 131 Harv. L. Rev. 685, 691 n.17 (2018):While the judicial power to enjoin and find in contempt an agency official is not seriously questioned, it is uncertain whether, as a constitutional matter, the President could be enjoined or held in contempt. In any event, there are ways for judges to indirectly control much presidential action. See, e.g., Franklin v. Massachusetts, 505 U.S. 788, 825-29 (1992) (Scalia, J., concurring) (arguing that there are no historical examples of courts enjoining the President in his official capacity and that such injunctions would violate the separation of powers, but conceding that injunctions usually are available against the subordinate officials who are necessary to carry out a President’s order); United States v. Nixon, 418 U.S. 683, 692 (1974) (“The issue whether a President can be cited for contempt could itself engender protracted litigation . . . .”); Jonathan R. Siegel, Suing the President: Nonstatutory Review Revisited, 97 Colum. L. Rev. 1612, 1678-79 (1997) (noting that a few district court orders have run against the President and that noninjunctive declarations of the law have in fact induced the President to conform to a court’s holding). ↑
45. 603 U.S. 593, 642 (2024) (holding that the President “may not be prosecuted for exercising his core constitutional powers, and he is entitled, at a minimum, to a presumptive immunity from prosecution for all his official acts”). ↑
46. See, e.g., cases cited at notes 19-20, supra. ↑
47. Trump v. IRS, 836 F.Supp.3d 1225, 1262 (S.D. Fla. 2026). The Eleventh Circuit subsequently denied the motion for stay pending appeal filed by Trump and other plaintiffs. ↑

The post Servant of the Law: The President as Chief Prosecutor appeared first on Just Security.

This trip is Rubio’s first visit to the region as secretary, and it’s a business trip as much as a diplomatic one. It builds on the earlier C5+1 Washington summit in 2025, when Secretary of Commerce Howard Lutnick announced $25 billion in commercial deals with countries in the region, with hundreds of billions more in potential future agreements. The United States seeks access to Uzbekistan’s vast supply of minerals and rare earths, along with other possible deals around aviation and agriculture. Uzbekistan, for its part, is clearly interested in expanding trade and commercial ties with the United States. Today, its foreign trade is mostly concentrated in China and Russia, giving Moscow and Beijing clear leverage over Tashkent. 

By forming closer ties with Washington, Uzbekistan seeks to diversify its trading partners. And there is one thing Uzbekistan wants specifically: the repeal of Jackson-Vanik trade restrictions, which suspend most-favored nation status for certain trade partners based on human rights concerns. They have applied to Uzbekistan since 1991, but only recently, as Uzbekistan has opened its economy and pursued trade diversification, has their repeal become a priority. Only the U.S. Congress can remove these restrictions, giving the United States leverage it has rarely had in Tashkent and that it cannot afford to waste. 

The Trump administration already demonstrated in Belarus how effective the White House can be in securing the release of political prisoners when it chooses to use its leverage. Over the course of a year, President Donald Trump’s special envoy for Belarus, John Coale, has helped secure the release of hundreds of political prisoners in exchange for targeted sanctions relief. In Uzbekistan, it would cost even less. Tashkent is actively seeking partnership with Washington rather than extracting concessions, giving the United States the chance to decide whether to grant the repeal outright or tie it to something of greater value. And the release of prisoners is not unprecedented in Uzbekistan: early in his presidency, Shavkat Mirziyoyev freed more than 35 unjustly imprisoned people, so he may be willing to negotiate this time.

An Easy Ask Before Doing Business

Today, dozens of prisoners of conscience from Karakalpakstan––a sovereign republic within Uzbekistan––are languishing in Uzbek prisons with no hope for justice. Four years ago, in July 2022, many of them took to the streets to peacefully oppose amendments to Uzbekistan’s constitution that would have removed Karakalpakstan’s sovereign status. Uzbekistan responded with deadly violence as security forces used lethal force against the demonstrators. Dozens of people were killed and nearly 300 were injured. Now, more than four years later, there has been no meaningful investigation into the actions of the Uzbek security forces and therefore no justice for the victims. Instead, Uzbekistan went after the protesters, imprisoning more than 60 individuals for their participation in the protests, many of whom are still behind bars today.

None have been punished more severely than Dauletmurat Tazhimuratov, a Karakalpak lawyer and journalist who gave a public speech condemning the amendments and calling for peaceful protests against them. For his role in the protests, Tazhimuratov received a 16-year sentence, the longest of any of the defendants, which has since been extended, bringing his total sentence to nearly 18 years. (Through our work at the Human Rights Foundation, we represented Tazhimuratov in his case before the United Nations Working Group on Arbitrary Detention and actively advocated for his release.)

While imprisoned, Tazhimuratov is suffering constant violence and abuse at the hands of prison officials. He is held in an overcrowded 4×5 meter cell and is only able to access bathing facilities about once every six weeks, so he is forced to bathe in the toilet area. Officials often deny him access to pen and paper, and they have limited visits from his family in an apparent attempt to cut off his contact with the outside world. Perhaps worst of all, high-ranking officials at the prison regularly beat and taunt him, trying to force him to disclaim his Karakalpak identity. One beating was so severe that he had blood in his urine for more than a week, but he was denied medical treatment. 

The U.N. Working Group on Arbitrary Detention found that Tazhimuratov’s detention violates international law, and multiple U.N. Special Rapporteurs have written to the government about his case. Recently, U.S. Representative James McGovern wrote to Uzbekistan’s foreign minister about Tazhimuratov’s case, calling for transparency and compliance with international human rights obligations. Despite all this, Uzbekistan has thus far ignored calls for his release and demands for better treatment in prison. But now is the perfect opportunity to demonstrate the power of direct U.S. leverage to deliver what years of international pressure have not.

Uzbekistan wants to diversify trade, expand its U.S. investments, and end its Jackson-Vanik designation. Washington has also signaled interest in deepening trade and investment ties with Uzbekistan, and talks are already underway to repeal Jackson-Vanik, with Rubio himself supporting legislative action to remove the designation. But the United States should not forfeit its leverage too cheaply. The United States presents itself as a bastion of freedom and this administration, despite other valid criticisms, has established a reputation for freeing not only Americans held abroad, but also non-citizens, such as Pastor Ezra Jin or the prisoners in Belarus. Washington should continue this pattern in Uzbekistan and not agree to repeal the trade restrictions without demanding freedom in return. This is the art of the deal.

Timing is everything. Before any agreements are signed or restrictions are lifted, Rubio and the business executives traveling with him must exercise their leverage and make it clear that commercial diplomacy with the United States comes with conditions. Uzbekistan cannot be allowed to continue arbitrarily detaining Karakalpaks for exercising their basic human rights, and the United States cannot passively accept that many of those arbitrarily detained are subjected to torment at the hands of government officials. 

If the United States is committed to closer cooperation with Uzbekistan, it cannot turn a blind eye to Uzbekistan’s forgotten. If the United States takes its commitment to freedom seriously, then Rubio must insist on justice for Dauletmurat Tazhimuratov, justice for Karakalpaks, and justice for all of Uzbekistan’s prisoners of conscience. 

The post To Secretary Rubio: Demand Freedom for Uzbekistan’s Political Prisoners appeared first on Just Security.

A curated weekday guide to major news and developments over the last 24 hours. Here’s today’s news:

IRAN WAR

Iran has managed to export some liquefied petroleum gas and jet fuel since July despite a U.S. naval blockade, shipping intelligence showed and sources told Reuters. Analysis by U.S. advocacy group United Against Nuclear Iran shows at least eight gas tankers carrying Iranian LPG have left the Gulf of Oman since mid-July. Feras Dalatey and Jonathan Saul report.

IRAN WAR – YEMEN

Yemeni health officials yesterday said the Iran-aligned Houthis struck Maashiq Palace in the southern city of Aden, the headquarters of Yemen’s internationally recognized government, killing five people and injuring at least 16. Yemeni Information Minister Moammar al-Eryani said Yemeni Prime Minister Shaya al-Zindani was not hurt. Houthi military spokesperson Yahya Saree said its fighters had bombed Saudi military operations headquarters in Aden, as part of the group’s declared policy of targeting “Saudi reinforcements wherever they were located.” Ismaeel Naar and Pranav Baskar report for the New York Times; Enas Alashray, Mohammed Ghobari, and Eman Abouhassira report for Reuters.

LEBANON

U.S. Secretary of State Marco Rubio during a call pressed Lebanese President Joseph Aoun not to return to Hezbollah explosives the group attempted to smuggle into Lebanon’s capital Beirut last week, sources told Axios. Rubio urged Aoun to “hold accountable all those who ordered, financed, and directed the shipment, and underscored that returning explosives to Hezbollah should be out of the question,” according to a U.S. State Department official. Rubio also raised concerns about Lebanon’s 1955 Boycott Law, which bans any contact with Israelis and told Aoun the law is “routinely used for politicized prosecutions that deter steps towards peace with Israel.” Barak Ravid reports. 

WEST BANK VIOLENCE

Some 80 Israeli settlers in face masks today burned olive trees and threw rocks at Palestinians at the West Bank village of Abu Falah’s annual harvest, injuring at least two people including a journalist, witnesses told Reuters. Israeli soldiers arrived at the scene and fired tear gas in the direction of Palestinians, and a Reuters photographer saw at least one armed settler fire a gun. Pesha Magid reports.

RUSSIA-UKRAINE WAR

The EU today is hosting Ukrainian President Volodymyr Zelenskyy in Brussels. Zelenskyy told the summit he does not see “signals from the Russian side that they are ready… to end this war.” Estonian Prime Minister Kristen Michal told Reuters that the EU would probably have to provide a new loan to Ukraine to help cover its budget gap for next year. Andrew Gray, Philip Blenkinsop, and Max Hunder report.

Russia yesterday publicly acknowledged a Ukrainian drone strike on a major Russian oil refinery in the Ryazan region and broadcast images of the damage on state media, breaking from usual practice to highlight an attack Trump branded a “disgrace.” The aftermath of an attack on a refinery was shown on TV for the first time since the start of the war. Reuters reports.

The order to limit SpaceX Starshield services used by Ukraine for battlefield communications and drone operations came from inside the White House, U.S. officials told Reuters. The officials said leaders in the U.S. Defense Department received a message that the SpaceX Starshield service for the Ukrainian military needed to be slowed, but did not specify who in the White House made the decision or who was responsible for implementing the order. Erin Banco and Gram Slattery report.

Ukraine has run out of missiles for the U.S.-made Patriot air defense system, a source told CNN yesterday. For months, Zelenskyy has repeatedly warned that Ukraine is running out of U.S.-made Patriot interceptor missiles, which are one of the few defensive weapons capable of knocking down a ballistic missile. The source said “almost all the ballistic missiles from the last strike were not intercepted due to the lack of the [interceptor] missiles.” Victoria Butenko and Lauren Kent report.

U.K. Defense Minister West Streeting yesterday said he and U.S. Defense Secretary Pete Hegseth welcomed the restoration of the Starlink service in Ukraine during talks. Streeting stressed the need to move forward in supporting Ukraine following renewed discord between the United States and Ukrainian leaders and said it was “not the first time we’ve seen a falling out between President Trump and President Zelenskyy.” Jasper Ward and David Brunnstrom report for Reuters.

OTHER GLOBAL DEVELOPMENTS

Bulgarian Defense Minister Dimitar Stoyanov today said Bulgaria has formally asked NATO to deploy combat aircraft and anti-drone systems to bolster its defenses after recent drone attacks on commercial shipping in the Black Sea. Stoyanov said he planned to meet his Romanian and Turkish counterparts soon to discuss coordinating action and a common approach to the situation in the Black Sea. Stoyan Nenov and Lili Bayer report for Reuters.

Two members of South Africa’s largest political party, the African National Congress, yesterday were fatally shot while campaigning. The attack is the second fatal shooting on party members this week, as on Tuesday four other members were killed as they were leaving a memorial service. The victims have not been identified but the ANC said they were not candidates and described them as volunteers. Zimasa Matiwane reports for the New York Times.

Ethiopian Prime Minister Abiy Ahmed today named a caretaker government to stabilize the northern Tigray region and prepare for an election within six months. Abiy’s office said the temporary administration will be in office for six months and Abiy justified the measure by saying Tigrayans can “attain the peace that has long eluded them, to secure their right to self-administration, and to finally enjoy the fruits of development.” George Obulutsa reports for Reuters.

Brazil’s Superior Electoral Court yesterday ruled that Meta must adopt measures to immediately secure equal treatment of the country’s two presidential candidates ahead of their Oct. 25 runoff. Brazilian President Luiz Inácio Lula da Silva filed a complaint that accused Meta of using biased algorithms and AI restrictions to favor opposition candidate Sen. Flávio Bolsonaro. The judge said Meta must pay a fine of $98,000 a day if it fails to comply. Mauricio Savarese reports for AP News.

U.S. AI DEVELOPMENTS

The Federal Aviation Administration is working to acquire an ownership stake in Air Space Intelligence, a private company that is key to the FAA’s efforts to modernize the United States’ air traffic control system, according to an internal FAA document reviewed by POLITICO. The FAA in June awarded the company a contract worth $875.9 million for an AI-powered tool to help air traffic controllers, and the AI platform’s initial launch in the Washington area sparked concern among Democrats in Congress who said it should be put on pause until Congress has a better understanding of how the tool is being deployed. Sam Ogozalek reports.

U.S. FOREIGN AFFAIRS

Trump yesterday said the United States is holding talks with Russia and China aimed at denuclearizing the countries. Trump said he expects China to “catch up” with the United States’ nuclear arsenal within the next five years, which he called a “shame, because I’d love to denuclearize if possible.” Jasper Ward reports for Reuters.

Canadian Prime Minister Mark Carney yesterday told a pro-Canada rally in Calgary, Alberta that the United States is trying to break Canada so it can “own us.” Alberta next Monday will vote on whether to remain a province of Canada or instruct the local government to begin the legal process toward a future referendum on separation. Carney said the referendum “comes at the worst possible time” as the world “has become more dangerous and divided.” Rob Giles reports for AP News.

Assistant Secretary of State for Intelligence and Research Michael Vance last week held talks with Myanmar’s government in Tokyo, sources told Reuters. The meeting involved more than a dozen U.S. officials, but Reuters could not establish who represented Myanmar. The State Department did not respond for comment on the latest talks but said it engages with all groups in Myanmar on issues affecting U.S. interests. Simon Lewis, Poppy McPherson, and John Geddie report.

U.S. IMMIGRATION DEVELOPMENTS

Microsoft Vice Chair and President Brad Smith on Wednesday told POLITICO that it was ironic the White House chose to target his company’s foreign workers on the same day it gave a medal to Microsoft CEO Satya Nadella, a former H-1B skilled visa holder from India. Smith said five of the six individuals who received medals from the White “had come to the United States as immigrants,” and that the Trump administration risks damaging the standard of living for Americans with restrictions on the visas. Brendan Bordelon reports.

The daughter of Jimenez Mendoza, a man who died in 2017 after falling from a cliff while he fled ICE agents near the southern border, on Wednesday told the New York Times that he was being depicted as living fugitive Jose Mendoza-Cervantes. Mendoza-Cervantes entered the vehicle driven by Miguel Parra Hernandez on Saturday in Houston, Texas, and Hernandez was later struck from behind with an SUV as he ran away from ICE agents on foot. Ms. Jimenez said her father used aliases, including the name ICE released as its target in the Houston operation. Court records showed Mendoza faced the same criminal charges that ICE attributed to Mendoza-Cervantes. Jazmine Ulloa reports.

The Justice Department yesterday sued California and a Los Angeles County official, seeking to temporarily block the state from counting mail ballots submitted by voters it suspects are noncitizens. The suit would apply to some 20,000 people in the state the Trump administration has cast as “likely noncitizens.” The suit wants to prevent the counting of votes “until their citizenship status can be determined.” Attorney General Todd Blanche said California had received ample warning and was given “an opportunity to self-correct its records.” Devlin Barrett and Laurel Rosenhall report for the New York Times.

The League of Women Voters, the Electronic Privacy Information Center, and the League of United Latin American Citizens yesterday sued the Department of Homeland Security, alleging the government has obtained voter files from at least 47 states and handled the data in a way that violates federal law. The lawsuit alleges the DHS’s secretive probe of the voter data disregards federal safeguards and flouts the privacy rights of millions of Americans. Ali Swenson reports for AP News.

A federal judge yesterday threw out a Trump administration lawsuit that sought to block New York City ‘sanctuary’ laws that limit cooperation with federal immigration enforcement. The judge said the administration could not challenge a city law barring police from detaining immigrants at the request of federal authorities because that practice is already prohibited by state law. Jack Queen reports for Reuters.

U.S. DOMESTIC DEVELOPMENTS

The FBI and U.S. Postal Service’s enforcement arm yesterday said they were investigating suspicious mailings, including some sent to state election officials and lawmakers’ offices. The FBI said some of the letters contained an unknown substance. Spokespeople for Senate Democratic Leader Chuck Schumer (NY), Sen. Alex Padilla (D-CA), and Sen. Chris Murphy (D-CT) all said the USPS intercepted letters containing a suspicious substance mailed to their offices. David Shepardson reports for Reuters.

SpaceX CEO Elon Musk and his America PAC political action committee yesterday said they agreed to settle two proposed class actions alleging Musk defrauded residents in closely contested states before the 2024 presidential election by promising voters a lottery-style daily $1 million giveaway. The filings did not include settlement terms. Mike Scarcella reports for Reuters.

White House Press Secretary Katie Zacharia yesterday said her California Bar exam results are “somewhat inconsequential” because she knew she wouldn’t need the bar in China. Zacharia said “I had enough credentials, which was my law degree, to be able to be in this practice,” and added that she was “not hired as White House counsel” when asked if she would weigh in on legal matters as press secretary. Julia Manchester reports for The Hill.

TRUMP ADMINISTRATION ACTION

Trump on social media yesterday said the execution of convicted Fort Hood shooter Nidal Hasan by firing squad will not be live-streamed. Trump said he thinks streaming the execution “might have been a strong deterrent for the World to witness how America deals with Terrorists and SCUM like Hasan but, many very good and well meaning people would rather not watch him die,” and added that the “execution may be viewed in person by the immediate families of the victims, if they so choose.” Avery Lotz reports for Axios.

Nvidia and Apple each committed $25 million for Trump’s planned White House ballroom, according to a list obtained by the Washington Post. Trump himself is not on the list. Jonathan O’Connell, Sarah Blaskey, and Aarton Schaffer report; Nandita Bose and Doina Chiacu report for Reuters.

Paolo Zampolli yesterday said he left his post as presidential special envoy in the State Department. Zampolli on social media suggested the exit was routine, but a U.S. official told POLITICO that Zampolli was fired. POLITICO earlier yesterday reported that Zampolli and the Deputy Secretary of State Christopher Landau had been embroiled in a monthslong email fight that included accusations of leaks to the media and threats of defamation. Chris Cameron reports for the New York Times; Daniel Lippman reports; Daniel Lippman reports.

State Department spokesperson Tommy Pigott yesterday said that former Homeland Security Secretary Kristi Noem is no longer serving as the Trump administration’s special envoy to the “Shield of the Americas” coalition. Pigott said Noem will continue “to be at the State Department as a senior adviser.” Ford McCracken reports for The Hill.

TRUMP ADMINISTRATION LITIGATION

A federal judge yesterday denied a request by crypto firm Gemini and the Commodity Futures Trading Commission to undo a court-approved settlement reached before Trump returned to office in Jan. 2025 which requires Gemini to pay $5 million for misleading federal regulators about its Bitcoin business. The judge said the facts of the case had not changed since Gemini, without admitting guilt, agreed to the fine. Sharon LaFraniere reports for the New York Times.

Nine former FBI agents and intelligence personnel yesterday sued FBI Director Kash Patel, Blanche, and their agencies, alleging they were unlawfully fired for working on the investigation into Trump’s handling of classified documents. The lawsuit alleges that the former employees were “unlawfully terminated not for any cause or any legitimate reason, but only to satisfy the President’s personal vendetta.” Josephine Walker reports for Axios.

A federal judge yesterday rejected a lawsuit from former C.I.A. Director John Brennan that sought to force the DOJ to preserve investigative records so that Brennan could challenge any eventual criminal case against him as a selective and malicious prosecution. The judge said Brennan’s concerns were too speculative at the moment and he had not proven that the DOJ was failing to preserve records relevant to a future malicious prosecution argument. Eric Tucker reports for AP News.

Did you miss this? Stay up-to-date with our Litigation Tracker: Legal Challenges to Trump Administration Actions.

If you enjoy listening, Just Security’s analytic articles are also available in audio form on the justsecurity.org website.

ICYMI: Yesterday on Just Security

Guantanamo Dispatches: Four Possible Obstacles to Jury Selection in the USS Cole Case at Guantanamo Bay

By Carol Rosenberg

The post Early Edition: October 16, 2026 appeared first on Just Security.

Here are some of the issues that could further delay the start of the Cole case, the first ever capital case at Guantanamo:

Competency Challenge: At the request of defense lawyers, Judge Fitzgerald ordered a Sanity Board to evaluate the capacity of the defendant, Abd al-Rahim al-Nashiri, to understand and cooperate with his lawyers. Earlier, a Pentagon-approved civilian forensic psychiatrist, who was hired as a defense consultant, spent a week with al-Nashiri at Guantanamo and found him not competent. Now, the Sanity Board is underway, with a court-appointed Army psychiatrist visiting the base earlier this month to conduct interviews, including with al-Nashiri but without his lawyers present. The Sanity Board psychiatrist also was reviewing prison records to provide an expert opinion and answer questions posed by the judge. The deadline for his report is Oct. 16.

Collateral Competency Challenge: On Oct. 15, lawyers for al-Nashiri asked a federal judge, Royce Lamberth at the U.S. District Court in Washington, D.C., to issue a temporary restraining order on the trial, and then an injunction until the competency question is fully considered at the Guantanamo level. “The military judge ordered a competency evaluation but left no time for a meaningful hearing on whether petitioner is competent,” said the petition by Allison Miller, al-Nashiri’s capital lawyer. “He has three times declined to stay the proceedings pending the inquiry.” Al-Nashiri’s lawyer also noted that certain detailed brain scans of the prisoner have not yet been done. 

Challenge to Trial Evidence: Al-Nashiri’s lawyers also have sought a stay on the trial at the U.S. Court of Appeals for the District of Columbia Circuit on a different issue: They say prosecutors provided the defense team with exculpatory evidence that could shift the blame for the Cole bombing onto another former prisoner. But there’s a twist: Defense lawyers say the evidence was obtained through the torture of the other prisoner in Afghanistan in 2002, and it would be unlawful to use it, even if the judge approved its use. The larger question is before the U.S. Court of Military Commissions Review, as explained in the filing, which was just released. The other prisoner, meanwhile, is out of reach of the court. He was transferred to Oman for rehabilitation in 2025.

Motion to Disqualify the Judge: Al-Nashiri’s lawyers are also asking Judge Fitzgerald to disqualify himself. They argue in a 29-page motion filed Oct. 8 that the judge exceeded his authority by involving himself in the mechanics of trial preparation and then became so locked into a trial start date that he set unfairly inflexible deadlines to make it work. They point to an ex-parte meeting Judge Fitzgerald had in Virginia with the director of the office of military commissions, Liam Apostol, to discuss transportation of jury pool members and other scheduling details. Apostol testified in a hearing on Oct. 15 that the meeting involved a slide presentation for routine planning; defense lawyers argued that the prosecution typically had that role.

The post Guantanamo Dispatches: Four Possible Obstacles to Jury Selection in the USS Cole Case at Guantanamo Bay appeared first on Just Security.

Related Just Security coverage: This Back-to-School Season, Create Bully-Bot Free Zones for Children and Teenagers by Leah A. Plunkett (Oct. 8, 2026)

You can read the documents in full at the links above and provided in full below:

Arizona

Arkansas

Idaho

Iowa

Kansas

Montana

Nebraska

West Virginia

The post Just the Docs: Eight States Sue OpenAI, Alleging It Misled the Public About ChatGPT’s Safety appeared first on Just Security.

This is a shift from what has long been a bipartisan position that American political, economic and security interests are advanced through anti-corruption efforts. Administration after administration has maintained – even if just as a goal – that countering corruption globally is important for creating a level playing field for U.S. businesses abroad, protecting the U.S. financial system from corrupt activity, and promoting political stability and reliability. Further, there has been an increasing recognition that corruption presents a range of national security risks. It opens space for terrorists and criminals to operate, betrays national sovereignty to adversary states seeking to exploit corruption, and props up authoritarian regimes, as former Senator Rubio once observed.

The United States should re-embrace the importance of anti-corruption in its foreign policy. As Coordinator on Global Anti-Corruption during the Biden administration, I worked across the government and with foreign partners to make some progress on this front, albeit imperfectly. We undertook internal reforms to make our policy and tools more coherent, thoughtful, and agile. We led dedicated initiatives to uncover and combat corruption, prevent corrupt actors from using the U.S. financial system, and shield U.S. businesses from corruption. I am skeptical that such a change could happen under the current administration; still, reflecting on the current anti-corruption landscape, I believe there are lessons we can draw on to improve our effectiveness and reinforce the value of countering corruption as a national security imperative in future ones. These lessons fall into four categories: we spread ourselves too thin, underinvested in accountability, moved too slowly when conditions changed, and did too little to address corruption at home. But each also points to a fix, which I lay out below. 

Lessons Learned

1. We were working hard and smart, but not smart enough.

The U.S. government’s anti-corruption enterprise spanned the gamut from supporting law enforcement operations to development support to public awareness campaigns. Offices throughout the U.S. government worked with foreign governments, individual NGOs, banks, companies, and international organizations like the United Nations to advance the agenda, consistent with an ambition to engage in a multidisciplinary anti-corruption effort.  

In practice, though, the U.S. government too often prioritized span over depth and reach over results. Spread across multiple agencies and multiple priorities, the U.S. government’s foreign assistance investment was not as impactful as it might have been. For years, bureaucratic territorialism, inertia, and complexity have at times driven redundancy and siloed planning, both in Washington, D.C. and at posts abroad. A lack of investment in shared frameworks and evaluation, particularly of impact, resulted in inadequate learning and disjointed planning processes. Ambition and idealism led to stretched resources, especially because it was hard to say “no” to objectively good ideas. And the absence of high-level political engagement, and an occasional reluctance to pursue anti-corruption actions that would ruffle diplomatic feathers, limited the effectiveness of the overall campaign. 

By the end of the Biden administration, the U.S. government had made some progress in addressing how to prioritize our work. At the State Department, we were working through a process to focus resources in specific countries to deal with the most salient problems. We also completed a multi-year effort to map out with greater precision the sorts of existing programs underway and use that common picture to identify redundancies and gaps in coverage. In some places, such as Ukraine, there was also sustained political will to prioritize anti-corruption and to resource it both financially and politically. Nevertheless, there was substantial room to improve. 

Today, there is no indication that the Trump administration has pursued any such improvements since coming into office. Indeed, the Office of the Coordinator on Global Anti-Corruption, which I led, ceased to exist in April 2025, and several other offices and agencies focused on countering corruption have lost staff and resources. 

2. We underprioritized accountability.

U.S. anti-corruption policy invested heavily in prevention and institution-building, but too often stopped short of ensuring that those who benefited from corruption actually received consequences. That gap was particularly striking because the demand for accountability was among the clearest messages we heard from the people confronting corruption firsthand. In several conversations with civil society, businesses, and even government partners in countries we visited, one message was clear: foreign populations want accountability for corruption, and they want U.S. support in achieving it. By accountability, we mean a proper rendering of consequences for corrupt action, whether within a foreign jurisdiction’s political or justice system or through external tools such as U.S. law enforcement and justice processes, financial sanctions, or visa restrictions. 

To advance accountability, the U.S. government worked across agencies to help reform foreign government institutions, to execute transnational law enforcement operations (including through organizations such as the International Anti-Corruption Coordination Center), and to deny access to the United States as a safe haven through use of sanctions and visa restriction tools. Some of these accountability steps were significant – such as the decision to impose sanctions on Horacio Cartes (former president of Paraguay) and Antal Rogan (senior advisor to former Prime Minister Victor Orban of Hungary) – while others were under the radar, working with partners to pursue their own investigation. (Of course, sanctions on both of these individuals were subsequently rescinded without any adequate explanation by the State Department other than that they were no longer seen as consistent with our foreign policy, making one question – at the very least – where fighting corruption fits into that policy.)

Still, accountability support requires more resources and commitment than we were prepared to push. We missed opportunities for foreign assistance investments and could have done more to work with foreign law enforcement and associated intel organizations. It was laborious to try to improve the prioritization of intelligence gathering and analysis on corruption, despite its manifest utility in supporting accountability efforts like sanctions, though some limited progress was made toward the end of the Biden administration.  

And, we could not focus sufficient sanctions resources on priority cases where reinforcement of existing sanctions, expansion to include enablers of corrupt individuals, and intensification of consequences on bad actors might have had an impact. Ironically, targeted financial sanctions and travel bans might be among the most effective measures that one could apply against corrupt individuals, as they deny such actors the benefits of their crimes. But, given limited resources and the many other priorities that existed in the U.S. government for sanctions use, the result was too few anti-corruption designations per year to cover key jurisdictions or problem areas.

Anti-corruption institutions are difficult to sustain when corrupt actors continue to operate with impunity. Prevention, transparency, and institutional reform remain essential, but they are more credible when paired with a realistic prospect of consequences. Underinvestment here is also a shame because, in addition to minimizing our anti-corruption impact, we also lost the ability to advance other priorities such as countering narcotics trafficking and terrorism that are often supported and facilitated by foreign corruption. Arguably, we’ve gone backwards in this regard as well now, with the effective derogation of the beneficial ownership database requirements of the Corporate Transparency Act. Now, instead of gathering information on who owns the companies operating in the United States to support law enforcement and deter criminals, we are reopening the door to illicit finance and associated corruption.

3. We were nowhere near as nimble as we should have been.

Anti-corruption conditions can change quickly but U.S. assistance and decision-making structures generally do not. When opportunities arose, crises struck, or reforms stalled, we often did not respond with anything like the flexibility or agility necessary. That flexibility matters because anti-corruption assistance is highly dependent on conditions on the ground. Meaningful reform generally requires credible domestic partners, and continuing to fund government institutions that lack the willingness to reform can waste resources or even reinforce the systems the United States is trying to change. Former Ambassador to Zambia and current Deputy Assistant Secretary of State for International Narcotics and Law Enforcement Michael Gonzales makes a strong argument that for any support to be effective, we need a committed partner on the other side. Continuing to fund government institutions that lack that commitment can waste resources or, worse, reinforce the very systems we are trying to change. The Biden administration recognized this in response to judicial corruption in El Salvador, for example, by pivoting funding from El Salvadoran government agencies to civil society.

But, it is frankly difficult to make those sorts of adjustments given the structural issues involved in U.S. budgeting and the policy or personal dynamics involved in shifting around resources. We must also seriously consider the risks of moral hazard and steward taxpayer dollars accordingly.

To the first point, moving congressionally-appropriated money around is often difficult by design and for good, constitutional reasons. The current administration is showing some of the perils that come from ignoring congressional appropriations, whether as it relates to foreign assistance appropriated or the misuse of taxpayer dollars altogether. But, the solution to a rampant, malfeasant abuse of congressional appropriations is not to make it impossible for law-abiding administrations to adjust funding as circumstances change.  

When there are windows of opportunity, reforms are not forthcoming, or money is being futilely spent, it is in the U.S. interest to be able to repurpose funds more quickly. After all, anti-corruption funding is scarce and there is intense competition for its use; spending U.S. taxpayer dollars in anti-corruption programs that have little chance of making a contribution is not only wasteful, it is also a distraction for those implementing and supervising these programs. Making these decisions can be particularly challenging in anti-corruption work because meaningful and durable improvements can take time, necessitating some patience in execution but speed in getting them started.

In the case of Guatemala, we can see the benefits of moving quickly but also the difficulties in doing so. When the Guatemalan government was growing mired in corruption and dismantling institutions like the International Commission against Impunity in Guatemala (known as CICIG for its Spanish acronym) under the presidencies of Jimmy Morales and Alejandro Giammattei, it took time to reposition funding and support to civil society actors in Guatemala. Then, when Bernardo Arevalo was elected, the United States was quick to take action to ensure that he would not be denied office. But, it was harder to surge resources into Guatemala to support the democratic transition and associated institutions. Of course, the Biden administration did try to do a lot to support President Arevalo upon his election in 2023. But, primarily given the long-term nature of other commitments to other countries made by the U.S. Department of State and USAID, it was nowhere near as much as it could have been.

Still, deciding to end support for a foreign government or even a country program focused on civil society actors brings tradeoffs and risks, too. The absence of engagement can increase the potential for security threats or internal instability with ripple effects. Further, such an admission of failure could be perceived as a surrender to corrupt actors, taking the wind out of the sails of champions and civil society organizations with real chances to achieve change; it could even expose those people to risk of retribution if the government involved believes the United States would look the other way. Yet, providing such funding with little prospect of near-term improvement also risks depriving other countries and their civil societies of valuable resources. 

4. Our own house is still too dirty.

U.S. shortcomings matter in two ways: weaknesses in our own financial and legal systems can enable corrupt actors abroad, while failures to address corruption and conflicts of interest at home undermine our credibility when we ask other governments to do so.

I often spoke with humility in fighting corruption given our own national issues with corruption and its enablers, across the political spectrum. I cannot imagine how difficult such a conversation would be today after the first 20 months of the Trump administration.

The United States itself has long provided avenues through which the proceeds of foreign corruption can be hidden, moved, or protected.

Even our positive words and actions – such as the Treasury Department’s important steps to enforce the Corporate Transparency Act (CTA) and other financial integrity measures – did not minimize many countries’ views that the United States contributes to an environment receptive to transnational corruption and its enablers. Our refusal to pass a version of the Enablers Act or properly resource the Financial Crimes Enforcement Network meant that countries often said that we were a source of their problems without being part of their solutions. And now, the Trump administration is willfully flouting its legal obligation to enforce the CTA, again undermining global anti-corruption efforts by allowing vulnerabilities for illicit finance by U.S. corporations and Limited Liability Corporations (LLCs) to persist.   

Other longstanding weaknesses in U.S. anti-corruption and ethics laws have also hurt our credibility abroad. Examples include the absence of senior executive ethics rules and enforcement, Congressional stock trading allowances, and campaign finance issues. As documented elsewhere in Just Security’s When the Guardrails Erode series, the United States has regressed on a range of anti-corruption standards under the Trump administration, and the ability to exploit these gaps is more visible.

Lessons Applied

To fix these issues, I suggest some potential solutions. This administration’s record makes it extraordinarily unlikely that such a commitment is plausible in the near-term. Thus, these recommendations are targeted at Congress and future administrations that might make a sincere effort at anti-corruption.

At least four things ought to be part of the solution:

1. Institutionalize regular strategic planning for anti-corruption and assign clearer roles and responsibilities to and within agencies.

The White House should be willing to make hard decisions about how to prioritize and oversee the whole-of-government strategic planning on anti-corruption, with close agency consultation, and then execution by agencies –– similar to the Biden administration’s National Strategy on Countering Corruption. But this process should include a detailed and firm delineation of priorities, roles, responsibilities, and action planning processes. This will make it easier to execute priorities globally and at a country level, help minimize overlaps, redundancy, and territorialism, and ultimately ensure that U.S. contributions are meaningful and not spread too thin.

Anti-corruption development support can and should remain available to facilitate institutional reform, equipment purchases, and the like. Still, it should be seen as what it is: a long-term investment that should only be made in countries where there is long-term potential and a genuine commitment by government actors. In countries where there is no such commitment, the priority investment must be in sanctions accountability measures and justice processes to break corrupt actors’ hold. In some cases, the focus should be on defending and supporting anti-corruption champions in-country, as the Biden administration did through various assistance programs and its Anti-Corruption Champions Award. 

To ensure effective implementation, the White House should remain engaged during implementation and gauge progress, including through more robust performance evaluation. Building on work in the Biden administration, it should also prioritize more robust in-country planning processes in priority countries. Finally, Congress should conduct proper oversight on the strategic approach and execution, direct funding authorities, and issue mandates as appropriate. Given the Trump administration’s choice to abandon or diminish many past bipartisan practices, this may be critical.

2. Put our money and political capital where our mouth is.

We should recognize anti-corruption as a relatively low-cost way to ameliorate a range of security and other foreign policy challenges, certainly as compared to the alternatives that can manifest when corruption helps to foster narcotics trafficking, criminal enterprise, and terrorism. Then, we should invest in it, with both money and political capital, sufficient to ensure adequate coverage and focus. As a baseline, we should have sustained, serious, and well-resourced support for sanctions and visa restrictions efforts and our own law enforcement focus on transnational corruption. This would include sufficient resources to the International Anti-Corruption Coordination Center (IACCC), U.S. staff covering sanctions and visa restrictions, financial crime enforcement and asset recovery, and intelligence collection. 

Resources alone will not be sufficient. Senior officials must also be willing to incur diplomatic costs when significant corruption cases implicate politically important partners or compete with other foreign-policy objectives. Where the investment is prudent, we should invest in foreign assistance to governments and civil society to advance in-country and transnational anti-corruption efforts, including promoting domestic accountability and prevention systems. This includes seizing on windows of opportunity like Guatemala or investing more in recent bright spots, such as Jordan, Moldova, or the Seychelles. For maximum impact, several changes have to be made: we cannot and should not be the lone anti-corruption enforcer, we have to make our investments and decisions more evidence- and results-based, and we have to message effectively that foreign commitment to anti-corruption is a U.S. priority equal to other national security requirements, not one to be abridged out of convenience. Finally, we should devote high-level political capital to diplomatic engagement, including when anti-corruption priorities are at odds with other priorities like critical minerals or security partnerships. 

3. Adopt more flexibility in authorities and implement quick decision-making.

Congress and the executive branch should develop lawful mechanisms that allow resources to be redirected when existing programs are no longer effective and surged when genuine windows for reform emerge. That could include contingency funding, greater transfer or reprogramming authority subject to congressional notification and oversight, and mechanisms for rapidly deploying technical assistance, investigative expertise, and other personnel.

Since multiple agencies are involved in the anti-corruption effort and use a wide array of tools, we need a way to assess the effectiveness of our interventions and to help decide next steps. We already have one tool for doing so: the report required under the Combating Global Corruption Act (CGCA). This bill, which passed into law in 2023, requires the State Department to evaluate and report to Congress annually until 2032 about progress being made in fighting corruption. A public report covers good progress; a private, classified report covers countries where there are continued problems and/or backsliding.  

Although State’s first released report was late and underwhelming (omitting countries like Sweden, for example, where anti-corruption efforts remain strong), the report remains a conceptual basis for making serious funding and support decisions. As part of its analysis and reporting process, State should identify those countries where more support would be beneficial or where continued efforts would waste U.S. resources. This should be accompanied by a request to redistribute any associated funding where a specific legislative requirement is in place or notification of an intent to repurpose funds where there is already some discretion. Importantly, State’s report should be developed with the support of other U.S. agencies in a process facilitated by the White House in its role as the coordinator of U.S. foreign anti-corruption policy. In cases where there are rapid changes on corruption issues, the White House, agency leadership, and embassies should run quick decision-making processes, building on past practice while innovating on how we collect and use evidence.

Congress should also invest in the sorts of tools and intelligence support available to target corrupt actors abroad. More flexibility in sanctions tools to decide whether asset freezes or travel bans are appropriate in all cases, and how to create conditions for sanctions removal in exchange for restitution by corrupt actors, would be especially welcome.

4. Take more action at home.  

The United States, across branches of government and at federal and state levels, must confront the ways we enable or permit transnational corruption here at home. We should pass the Enablers Act to make it harder for service providers in the United States to help corrupt actors loot their countries and exploit our friends. We should properly enforce the CTA, the Foreign Corrupt Practices Act, the Foreign Extortion Prevention Act, and the CGCA. We should recreate and invest resources into the Department of Justice’s anti-kleptocracy efforts and similar structures elsewhere in government. If there are good faith, practical and serious suggestions from the private sector and others to make these tools more efficient, they should be incorporated. But the imperative should be on cleaning up the mess we have created.

To practice what we preach, we must reverse backsliding and recommit to countering corruption at home, consistent with standards in the CGCA that we have established for foreign countries. Examples include: ensuring proper resourcing for domestic anti-corruption investigations, prosecutions, and government oversight; undertaking campaign finance and conflicts of interest reforms, including for the President; limiting Congressional stock trading; responding to the indications of rampant participation in prediction markets by U.S. officials with insider information; implementing anti-corruption safeguards for cryptocurrency; defending civil service protections, and many others.

Conclusion

Apathy in confronting corruption is common. It is seen as too hard a problem, one that has long been with humanity and is impossible to vanquish. That is likely true in an absolute sense, but as with other behavior we consider harmful to society, like theft, we should still try to prevent or combat it as much as possible. Even facing challenging global realities, we can remain true to such principles. The pernicious costs of corruption — to security, economic prosperity, political order, and beyond — make it well worth the effort. 

Author’s Note: Special thanks to Schuyler Miller, former Policy Advisor to U.S. State Department Coordinator on Global Anti-Corruption Richard Nephew, for his contributions to this article earlier this year.

The post Lessons from U.S. Global Anti-Corruption Efforts and Opportunities Ahead appeared first on Just Security.

Models such as Mythos raise questions about access to frontier models and institutional capacity: who can act directly on what these systems discover, and who must rely on others to interpret, prioritize, and fix their vulnerabilities? Hence, Anthropic’s decision to withhold Mythos from the vast majority of users foregrounds how access restrictions, which seek to delimit risk, ironically raise second-order risks. This situation introduces the risk of concentrating defensive knowledge and capabilities, potentially amplifying cascading harms when such power is held by a narrow set of actors in the absence of robust AI governance frameworks.

This essay takes this conundrum as its analytical point of departure. As discussed below, this approach to Mythos offers an aperture to examine digital sovereignty in an era when control over advanced AI increasingly shapes who can discover, defend against, exploit, and govern critical digital infrastructure.

Digital Sovereignty and Interdependence

This advancement comes amid intensifying debates over digital sovereignty. For several years, digital sovereignty—a policy agenda predicated on a government’s capacity to develop, deploy, and govern digital technologies on terms it can meaningfully shape—has stressed data localization, compute, platform regulation, and institutional capacity-building. These discourses and ambitions unfurl within an asymmetric context in which much of the digital infrastructure that underpins Information and Communication Technology (ICT) sectors across the world, and particularly in the Global South, remains financially and technically dependent on U.S. and Chinese providers.

It is important to note that for many governments, the pursuit of digital sovereignty is inseparable from mounting perturbations over the weaponization of interdependence, which has revealed chokepoints that index both the successes of globalization and its vulnerabilities. By invoking digital sovereignty, governments seek to restructure domestic ICT sectors, reduce reliance on externally governed tools viewed as possible strategic liabilities, and, most saliently, ensure that the control of digital infrastructure allows emerging technologies to reflect local priorities, values, and development needs.

In this light, Mythos complicates calls for disentanglement, industrial policy, and national champions. For one thing, localizing data and infrastructure is believed to advance these pursuits of digital sovereignty. However, digitization itself hinges on interdependence: chips, cloud systems, vendor updates, and other externally supplied components and services. Accordingly, even domestically owned infrastructure may remain dependent on external systems of maintenance, diagnosis, and repair. Mythos makes this condition more visible. In an AI-mediated cybersecurity environment, digital sovereignty will rely not only on who owns data or infrastructure but also on how dependency is negotiated with allies and rivals. More fundamentally, it will increasingly turn on who controls frontier models—often developed or governed by leading U.S. and Chinese labs—capable of discovering, securing, exploiting, and helping remediate vulnerabilities embedded across expansive digital infrastructures.

Digital Chokepoints

Project Glasswing, Anthropic’s controlled-access cybersecurity initiative, brings these challenges of digital sovereignty and dependency into sharp view. It provides selected companies and institutions with early access to Mythos so they can identify and remediate vulnerabilities before comparable capabilities become more widely available. According to Anthropic, participating organizations can use Mythos Preview for defensive work, including vulnerability detection and patching automation.

Project Glasswing is, in many respects, a reasonable response to proliferation risk. If frontier models can accelerate vulnerability discovery, then software developers, cloud providers, financial institutions, and critical-infrastructure operators have a legitimate interest in gaining time to strengthen their immunity. A defensive head start may reduce the likelihood that newly discovered vulnerabilities are exploited before safeguards can be developed. Yet this move does not address the deeper governance questions: who is (and is not) granted access to these capabilities, and according to what technical, commercial, institutional, or geopolitical criteria? 

Limited public transparency about the criteria for granting, prioritizing, or withholding access further reveals the challenges of dependency. Governments seeking to diminish reliance on externally governed digital systems may acquire greater cybersecurity capacity only through frontier models provided by firms they neither own nor govern. In this sense, advanced cybersecurity AI becomes a chokepoint: a strategically consequential capability through which access to defensive preparedness—and the technical and temporal advantage it can confer—can be selectively allocated.

Hence, Anthropic is doing more than selling a product. They are helping to determine when such capabilities are released, which actors receive access, under what conditions, and at what cost.

Leverage and Discretion

This chokepoint leverage raises critical matters. Cybersecurity has long been treated as a key modality of state sovereignty: governments defend networks, protect critical infrastructure, regulate software security, and, in some cases, stockpile vulnerabilities for intelligence or military purposes. Mythos complicates this arrangement. A leading AI laboratory may control a technology capable of discovering vulnerabilities more quickly than many state agencies, security researchers, cybersecurity firms, or criminal actors. Used responsibly and defensively, such a model could strengthen institutions and public safety. Used offensively, it could lower the barriers to sophisticated cyberattacks. Selectively withheld, even under reasonable concerns about proliferation, it could deepen asymmetries between those granted the capacity to protect themselves and those left comparatively exposed.

This challenge is especially acute in emerging markets, where banks, hospitals, power grids, mobile-money platforms, electoral databases, and government networks often operate through layered combinations of foreign hardware, proprietary software, cloud services, open-source dependencies, and outsourced maintenance. These arrangements already distribute technical authority unevenly: domestic institutions may use and regulate critical systems without possessing the expertise, source code, software inventories, dependency maps, institutional depth, or infrastructural control needed to secure them independently. 

Kenya’s experience with its surveillance system illustrates this point. According to a government audit, local personnel had not fostered the technical expertise required to operate and maintain the Chinese-supplied system independently, leaving key functions dependent on the contractor; even the control-room interface was still in Chinese. Even more alarming, the audit identified weak access-control protocols that increased the risk of unauthorized entry, malicious code being introduced onto the server, and the misuse of the surveillance technology by unauthorized users.

The stakes of this chokepoint reveal that cybersecurity tools operate as modalities of power. If the most advanced defensive tools remain concentrated among a small number of frontier AI firms, allied governments, and preferred partners, cybersecurity capacity may become another instrument through which dependence is governed. States and institutions outside trusted circles may struggle to anticipate vulnerabilities and become increasingly reliant on external actors during crises. What appears as responsible risk management may also contribute to an asymmetry of preparedness, allowing powerful firms and government partners to influence the security posture of others by controlling capabilities they need but cannot easily reproduce.

The Challenge of Inclusion

Exclusion from trusted circles is not the only problem. Inclusion can also become a source of dependency. This dynamic is not entirely new. Earlier waves of digital transformation revealed a similar pattern: governments have often expanded their administrative and security capacities through U.S. systems they do not fully command—from Nvidia’s dominance in advanced AI chips to the hegemony of the U.S. dollar. These arrangements can produce real capacity. They allow governments to connect with citizens, monitor threats, digitalize public services, and participate in global digital markets. But they also embed state action within external architectures of finance, maintenance, standards, upgrades, and technical control. 

In turn, Mythos extends this dynamic into the cybersecurity layer. It demonstrates that dependency can be produced not only through exclusion from advanced technologies but also through inclusion. Selected actors may gain access to powerful defensive capabilities, but the conditions of that provision may remain controlled by the model provider. Governments and firms become more capable, yet simultaneously more dependent on proprietary AI they neither own nor govern. Their security may improve, but the terms, duration, pricing, continuity, and distribution of that security are determined elsewhere. What they hold is not sovereign digital capacity but leased capacity.

This double bind suggests that countries need a coherent strategy to manage both the risks and opportunities of digital dependency. The aim cannot be to escape interdependence entirely. After all, even technological leaders such as China and the United States do not control the full digital stack. Rather, countries must negotiate from within their structural dependencies, delimit and mitigate their vulnerabilities, and use the opportunities afforded by building upon foreign platforms. That is what digital sovereignty should mean in practice: not the opiate of an untethered sovereignty, but the capacity to build leverage from within dependence.

Conclusion

The example of Mythos, then, clarifies the politics of digital sovereignty in an age of AI-enabled cybersecurity. The central problem is not simply whether governments have access to frontier models, nor whether provision should be unrestricted. Some limits may be necessary when systems can accelerate both vulnerability discovery and exploitation. The deeper question is how such limits are governed, who sets their terms, and how their benefits and burdens are distributed. When provision of cybersecurity AI is controlled by a small number of frontier labs and their state partners, defensive capacity itself becomes a chokepoint: a strategic capability that can be granted, delayed, priced, conditioned, or denied.

This matters because calls for digital sovereignty are, in part, responses to the weaponization of interdependence. Yet, the infrastructures through which AI operates cannot simply be disentangled from global circuits of chips, critical minerals, and cloud services. And so, the task is not to reduce digital sovereignty to a project of circumvention. Rather, it is to ask how interdependence is organized, who controls its chokepoints, and whether states and institutions can negotiate dependence in ways that strengthen mutual security and prosperity rather than reproduce vulnerability.

Concretely, this may mean that governments within trusted circles need to identify where critical dependencies actually reside and distinguish productive interdependence from vulnerabilities that structurally impede their capabilities. Procurement should then become a site of capability-building rather than outsourcing: governments can require greater transparency over software dependencies, local technical training, and viable exit or substitution pathways. Even so, these measures manage the form dependency takes rather than ending it. Frontier capability cannot simply be procured on better terms. Access ultimately remains something the provider can grant, condition, or withhold. However, for those outside trusted circles, the options are more limited. One avenue is to advocate through multilateral and multistakeholder forums—including the United Nations Global Mechanism on ICT Security and the Global Forum on Cyber Expertise—for mechanisms capable of governing the terms on which advanced cybersecurity capabilities are distributed, even though existing institutions do not directly govern the allocation of proprietary frontier capabilities by private firms. This limitation makes investment in domestic and regional technical capacity even more salient, including the pooling of capabilities—such as cybersecurity expertise and incident-response capacity—where national scale is simply insufficient.

In a word, the challenge is not exclusion, autarkic capacity-building, or concession to dependency, but rather the terms of interdependence. Mythos is not only a cybersecurity story. It is an analytical aperture into a wider geopolitical settlement in which the power to secure infrastructure—and the power to withhold that security—becomes central to the meaning of digital sovereignty itself.

The post The Mythos of Digital Sovereignty appeared first on Just Security.

A curated weekday guide to major news and developments over the last 24 hours. Here’s today’s news:

IRAN WAR

President Trump at a rally in Montana yesterday reiterated that he wasn’t sure who U.S. officials were negotiating with in efforts to end the Iran war. Trump said Iran doesn’t have a navy, airforce, or army left, and that he “should have probably left their leaders because I don’t know who the hell we’re negotiating with.” Abigail Walker reports for The Hill.

The U.S. military has developed new strike plans that include potentially using the “mother of all bombs” to target Iranian energy infrastructure and nuclear sites should Trump give the order to resume combat operations, U.S. officials told CNN. The use of the GBU-43/B Massive Ordnance Air Blast Bomb, the largest non-nuclear U.S. bomb, for the first time would mark a significant escalation in the war. Zachary Cohen, Haley Britzky, and Kevin Liptak report.

IRAN WAR- OTHER DEVELOPMENTS

The Maritime and Port Authority of Singapore and U.N. International Maritime Organization yesterday reported two incidents of ships sinking near the Strait of Hormuz. U.N. IMO reported on the sinking of an Antigua and Barbuda-flagged ship that sank close to the strait weeks after being struck by a projectile on Sep. 23. One crew member was killed and the rest of the crew was evacuated straight after the strike. Separately, MPA reported on a Singapore-flagged container vessel that sank while sailing through the strait, and confirmed all 21 crew members were safely evacuated. Jonathan Saul and Renee Maltezou report for Reuters.

LEBANON

U.N. Undersecretary-General for Peace Operations Jean Pierre Lacroix yesterday told AP News that Lebanese authorities “were unanimous in telling us they want to keep a U.N. presence” in Lebanon. The U.N. Interim Force in Lebanon monitors and defuses tensions between Israeli and Lebanese troops on the ground, and is due to end its mission soon despite a push by Lebanese officials for a last-minute extension of UNIFIL’s mandate. Abby Sewell reports.

ISRAEL-HAMAS WAR

A U.S. federal judge on Tuesday temporarily blocked the Trump administration from penalizing two U.S.-based organizations, Democracy for the Arab World (DAWN) and Taxpayer Alliance Againat Genocide, who support the International Criminal Court’s investigations into alleged war crimes in Gaza. The judge ruled that Trump’s executive order placing sanctions on the ICC violated the companies’ right to free speech under the First Amendment. Francesca Regalado reports for the New York Times.

ISRAEL

Israeli Prime Minister Benjamin Netanyahu yesterday said he was not in favor of a preemptive strike against Iran before Israel’s elections on Oct. 27 but also warned Iran and its proxies against striking Israel. Netanyahu said that strikes would not be “convenient for us, it’s no good.” Maayan Lubell and Lina Obeid report for Reuters.

Human rights groups Physicians for Human Rights Israel and the Public Committee Against Torture in Israel have reported a surge of mostly Palestinian inmates being shot by guards with rubber bullets at close range, with some suffering severe injuries, according to PHRI files reviewed by AP News. Of the 22 cases of inmates being shot by rubber bullets, 21 were Palestinian, and nearly all the shootings have occurred since last fall. The groups sent a letter to the Israel Prison Service in August urging it to end behavior they called “widespread” and “unlawful.” Sam Mednick reports.

A report by Canadian watchdog group Citizen Lab and another by U.S. advertising security firm Confiant today exposed that Israeli digital advertising agency Discover-Tech spied on internet users through a network of online ads it used to track targets and help hack their devices. Discover-Tech’s services gave Israeli surveillance firms access to advertising platforms run by Google and other search engines. Raphael Satter reports for Reuters.

RUSSIA-UKRAINE WAR

Ukrainian President Volodymyr Zelenskyy told the Financial Times in an interview published today that Ukraine will resume strikes on Russia’s oil refineries, and expressed anger over a “dangerous” withdrawal of U.S. intelligence and simultaneous weakening of Starlink satellite internet service at a time of intense Russian attacks. Zelenskyy said Ukraine paused energy strikes for three days to comply with U.S. wishes while Russia continued to attack Ukrainian energy infrastructure and kill civilians. Zelenskyy also said Trump had an instinctive affinity for autocrats and the super wealthy. Alec Russell and Ben Hall report.

Russia’s Ryazan region Gov. Pavel Malkov today said a Ukrainian drone attack caused a fire to break out at an oil refinery. The attack was the first by Ukraine since Trump announced an energy ceasefire last Sunday which neither side confirmed. Reuters reports.

Ukraine’s Kirovohrad region Gov. Andriy Raikovych today said the central region was hit by the largest Russian attack on its territory during the war overnight, killing three people and wounding another six. Kirovohrad said the “enemy deliberately hit critical infrastructure, transport enterprises, logistics facilities, and residential homes.” Max Hunder reports for Reuters.

Ukraine’s military today said it had damaged six Russian military aircraft in an overnight strike on the Borisglebsk airfield in Russia’s Voronezh region. Lina Golovnya and Max Hunder report for Reuters.

SpaceX CEO Elon Musk yesterday suggested that the U.S. government may be behind restrictions on the speed at which some of Ukraine’s weapons can travel and stay connected to SpaceX’s Starlink internet network. Musk said that Starshield, the military-grade sister service to Starlink, is controlled by the U.S. government, and “any limits on speed or usage are decided” by the U.S. military and Trump. Tiffany Wertheimer reports for BBC News.

Former Ukrainian Defense Minister Mykhailo Fedorov yesterday said SpaceX was working to restore satellite speeds in Ukraine and thanked Musk for his attention to the issue. Fedorov said he hoped the company would next “remove Starshield’s speed limit entirely” and that the move is “now critical for Ukraine’s defense.” Sarah Davis reports for The Hill.

Russia has nearly finished building a new base for launching jet-powered drones in just six weeks, according to defense experts and analysis of satellite imagery by Reuters. Tom Balmforth and Edward Carron report.

Russia’s FSB security service today said it had detained a Swedish man it suspects of planning attacks and sabotage on behalf of Ukraine after entering Russia on the pretext of joining the Russian army. The Swedish Foreign Ministry said it was aware of reports but was not currently involved in the case. Reuters reports.

German Chancellor Friedrich Merz today said frozen Russian assets in Europe should stay under the region’s control, following a report that the United States may seek supervision over some of the assets. Merz said it is Europe’s “business and we are deciding over it.” Ludwig Burger, Alexander Ratz, and Mrinmay Dey report for Reuters.

OTHER GLOBAL DEVELOPMENTS

Romanian officials today said two chemical tankers caught fire overnight inside Romanian territorial waters in the Black Sea after Russian drones strayed into Romanian airspace. Rescuers were searching for seven missing crew today from one of the tankers. Romanian officials stopped short of explicitly linking the fires onboard the ships to the Russian drones. Luiza Ilie reports for Reuters.

Brazilian President Luiz Inácio Lula da Silva and independent watchdog groups yesterday accused Meta and its social media platforms of giving his presidential rival Flávio Bolsonaro favorable treatment. The groups monitoring social media said search results omit progressive activists, videos critical of Bolsonaro are taken down without explanation, and posts praising Lula reach fewer people. Lula’s campaign filed a complaint with Brazil’s electoral court on Saturday demanding fair treatment online. Ana Ionova reports for the New York Times.

The New Yalu River Bridge between China and North Korea opened today. Chinese Premier Li Qiang said the crossing would become a new landmark symbolizing the two nations’ traditional friendship. AP News reports.

CLIMATE DEVELOPMENTS

A joint study by World Weather Attribution and Climate Central published today said that climate change and El Niño combined to create last month’s Hurricane Polo, which was the longest lasting Category 5 storm ever in the Pacific. Climate change made the conditions that fueled the hurricane 23 times more likely, while El Niño made it 14 times more likely. Seth Borenstein reports for AP News.

A study by the Danish Meteorological Institute published yesterday said an Atlantic Ocean current system weakened by global warming could recover over time if the climate cools. The study found no sign there was an irreversible tipping point for the Atlantic Meridional Overturning Circulation as suggested by previous studies, though the effects of the changes could last for centuries. Agnieszka Olenska and Louise Rasmussen report for Reuters.

U.S. AI DEVELOPMENTS

Saudi national and former OpenAI employee Abdulrahman Alfozan, who was this week charged with planting a secret recording device in an OpenAI executive conference room, last month appeared on an Arabic podcast interview and said he and other Saudi tech professionals in Silicon Valley had formed a group that advised Saudi organizations. The Saudi organizations included the kingdom’s insurance agency and RAED Ventures, a venture capital firm which names Alfozan as a venture partner on its website. Emmy Martin reports for the New York Times.

Eight states yesterday sued OpenAI, alleging the company has been misleading users about the safety of ChatGPT. The suit alleges the company’s failure to provide safeguards has encouraged suicide, self-harm, and violence among its users, including teenagers. Casey He and Elleiana Green report for POLITICO.

U.S. FOREIGN AFFAIRS

Former Venezuelan President Nicolás Maduro and his wife Cilia Flores are due in court today to respond to a new indictment accusing them of conspiring to torture political opponents. The additional charges were brought last Thursday and significantly expanded the criminal case against the pair. Their lawyers said they deny the torture charges. Luc Cohen reports for Reuters.

U.S. IMMIGRATION DEVELOPMENTS.

A Government Accountability Office report publicly released yesterday said federal immigration officials need to ensure body cameras worn by officers are uniformly activated when needed. ICE rejected the recommendation and said that using cameras that automatically activate when an officer pulls their weapon or activates a car siren could risk privacy concerns. Rebecca Santana reports for AP News.

Federal authorities investigating House Rep. Ilhan Omar (D-MN) seized her phone and have issued more than a dozen grand jury subpoenas, including to some of her relatives, a source told AP News. The subpoenas were issued for a grand jury collecting evidence in Minnesota as federal authorities probe potential fraud in Omar’s immigration paperwork. Alanna Durkin Richer reports.

A federal magistrate judge ultimately limited how the DOJ could access Omar’s documents in their raid on her house on Monday because of her role as a member of Congress, sources told CNN. Omar has been under criminal investigations for years, including before Trump’s new term, but Trump’s administration has pushed further with the belief she has repeated false statements when filing government documents such as passports. That is what led to the search on Monday, according to the sources. Katelyn Polantz, Evan Perez, and Jake Tapper report.

The lawyers of Oscar Belgal, the man who was shot by ICE agents in New York, yesterday said the agent who shot him never identified himself before he ran up to the car, smashed one of its windows, then fired several shots into the vehicle. The account by the lawyers sharply contrasted the account of events outlined by Department of Homeland Security officials, who said officers opened fire only after Belgal smashed his car into two government vehicles and “accelerated toward other officers.” Philip Marcelo reports for AP News.

California prosecutors on Tuesday filed felony assault charges against ICE agent David Gonzalez III who last Friday was found sleeping outside a home in Fresno at 3am and was shot by the homeowner during a scuffle. Gonzalez is charged with assault likely to produce great bodily injury, battery, elder abuse, and vandalism. The police said Gonzalez appeared “intoxicated and agitated” when the 72-year-old homeowner went to wake him, and Gonzalez assaulted the man and his daughter. Orlando Mayorquín reports for the New York Times.

A federal Court of Appeals yesterday ruled against a California law requiring ICE agents to show visible identification. The ruling stated that the federal government showed that the No Vigilantes Act violates the Supremacy Clause of the Constitution which “prohibits States from directly regulating federal operations even when regulating state operations in the same manner.” Ryan Mancini reports for The Hill.

U.S. DOMESTIC DEVELOPMENTS

The Justice Department yesterday threatened emergency court action unless California officials investigate potential ineligible voters within 48 hours and set aside ballots from anyone whose citizenship they cannot verify, according to two letters sent to state officials from the DOJ seen by Axios. Acting Deputy Attorney General R. Trent McCotter’s letter ended by saying that if California “cannot implement this process by then, given the ongoing federal election the United States will have no choice but to seek emergency relief.” Brittany Gibson and Avery Lotz report.

The FBI yesterday confirmed FBI Director Kash Patel was in Virginia for a law enforcement-related event and denied a report that he had traveled to Russia. MS NOW, citing two unnamed sources, earlier reported Patel had landed in Russia for a two-day meeting. Video footage from an ABC News’ affiliate in Virginia showed Patel at an event in Lynchburg, Virginia. Susan Heavey and Ryan Patrick Jones report for Reuters.

Just over half of major party candidates in the 13 most competitive midterm election states told Reuters they will accept election results regardless of the outcome, while some signaled a degree of uncertainty. David Morgan reports.

TRUMP ADMINISTRATION ACTION

Trump at a rally in Arizona yesterday said government sources who reveal information about military plans should be shot. Zolan Kanno-Youngs reports for the New York Times.

The Campaign for Accountability yesterday filed a complaint with the State Bar of California asking for an investigation into White House press secretary Katie Zacharia over her claims that she is a lawyer after reports surfaced that she is not licensed to practice law in any state. The complaint notes that if the reporting is accurate that Zacharia has “repeatedly held herself out to the public as entitled to practice law without a license to do so,” then it’s an act that California law treats as a crime. Mallory Wilson reports for The Hill.

Trump has directed his White House counsel to investigate claims brought forward on Friday by former Trump attorney Alina Habba, who alleged that former Kennedy Center leader Ric Grenell failed to deposit a nearly $6 million donation that her husband Turhan Mildon made to the center a year ago, sources told Axios. Grenell denied the accusations as “absurd” and a “lie.” Marc Caputo reports.

TRUMP ADMINISTRATION LITIGATION

The Supreme Court yesterday sided with the Trump administration and overturned a lower appeals court order to the Federal Communications Commission to finalize its rules outlining who can benefit from low rates for political advertising. The Supreme Court said the lower court’s decision was “inappropriate under the circumstances” and added that federal law “does not obligate the commission to act within a particular time frame.” The emergency order was unsigned and did not give a vote count. Abbie VanSickle reports for the New York Times.

Citizens for Responsibility and Ethics in Washington yesterday filed a lawsuit challenging the legality of Jared Kushner’s role in the Trump administration. The suit “seeks to bar Kushner from occupying his government position and exercising government authority in violation of the Appointments Clause.” The Appointments Clause lets the president appoint “principal” officers that the Senate later confirms. Kushner has not been confirmed by the Senate for his government work. Tara Sutter reports for The Hill.

Did you miss this? Stay up-to-date with our Litigation Tracker: Legal Challenges to Trump Administration Actions.

If you enjoy listening, Just Security’s analytic articles are also available in audio form on the justsecurity.org website.

ICYMI: Yesterday on Just Security

De-transition: Northern Ireland and the Ongoing Challenges of Keeping the Peace

By Fionnuala Ní Aoláin KC (Hons)

The post Early Edition: October 15, 2026 appeared first on Just Security.

It was all shockingly familiar for those of us who lived through “the Troubles” that stretched across three decades from 1969 to the Good Friday/Belfast Agreement of 1998. The crisis this month exposed some fundamental truths about the post-conflict transition in Northern Ireland, and it highlights the harsh reality that transitions from conflict are long and fragile, and we are often just a breath away from triggering a return to instability and violence.

For many years, colleagues outside of Northern Ireland would often remark caustically that I was too pessimistic about the success of the Northern Ireland transition. After all, our 30-year-plus deadly conflict had ended, political parties across the political divide had signed the Good Friday/Belfast Peace Agreement, weapons had been successfully decommissioned, and new democratic institutions were up and running. All of that is true. But living in Northern Ireland, close to the rawness of wounds, the ambivalence of the political settlement, and the inability to “deal with the past,” I knew we were living in a present that was still unsettled. The real question is just how unsettled?

This past couple of weeks has revealed the unresolved ambivalence of our peace. The challenge of translating the words of the peace agreement into operative practice on the ground presents a persistent test. For three decades, the fragility of post-conflict governance institutions – the legislative Assembly, for example — was displayed by their stop-and-start operation, politics regularly crashing on the shores of sectarian pressure. Non-state armed groups continue to pose a security threat; children’s education in Northern Ireland remains separate; and working-class communities continue to struggle economically. While conflict violence between (largely) male combatants in public ended, and while many communities have been able to live lives free of violence and fear, the necessary social and political movement towards shared space, forgiveness, truth, justice and reconciliation remains nascent. 

Slow and Winding Post-Conflict Progress

Thus, my assessment on the health of the peace process was always hesitant, though essentially positive. I soberly believed that this kind of stop-and-start progression, this laborious winding towards co-existence was, in fact, progress, after decades of conflict. Northern Ireland demonstrated that even in optimal post-conflict settings, progress would be steady but slow, uneven with backsliding, and that the legal commitments in the peace agreement would hold: to use “exclusively democratic and peaceful means of resolving differences on political issues,” and to maintain “our opposition to any use or threat of force by others for any political purpose, whether in regard to this agreement or otherwise.” 

After years of working in the ebb and flow of the transition in Northern Ireland as well as in conflict zones as a United Nations Special Rapporteur and advising governments on ways to end violence, I was wary of quick fix (and one-sided) solutions for ending conflict, and always cynical about the motives of mediators, negotiators, and bargains struck in other parts of the world that promised fast and happy endings in the short run. In Northern Ireland, we knew better.

Notwithstanding those standing caveats, the past few weeks have exposed deep imperfections in post-conflict institutions and processes designed to manage impasse and political accommodation. Tellingly, the Garvaghy Road dispute demonstrates the power of the spoiler in a peace process, individuals whose fundamental goal is to hinder, delay, or destroy the peace settlement. There are spoilers on both sides of the sectarian divide in Northern Ireland — those in the leadership of both Unionist (favoring Northern Ireland remaining part of the U.K.) and Nationalist (favoring unification with Ireland) political parties who gain more from stirring up sectarian sentiment than they do from stepping up and making compromise. But with elections pending and sectarianism selling better than solving our public services and budget crisis, the leadership of Sinn Fein and the Democratic Unionist Party have opted for the allure of the echo chamber and the comfortable silo. Non-sectarian leadership from the two dominant political parties is in low supply, and its articulation largely comes from smaller parties trying to hold the political middle.

The recent standoff also demonstrated another obvious truth: that politicians and some communities forget the cost of the conflict all too quickly. In Northern Ireland, deaths and conflict-related injuries have left profound intergenerational trauma, much of which remains unresolved because of the failure to embed a truth-recovery process in the peace deal. At least 20 deaths were related to standoffs over Orange Order marching in the 1990s, and it was the grim murder of three young boys in a petrol bomb attack in 1998 that finally forced retreat and compromise. 

Conflict Nostalgia

But that was, again, almost three decades ago. Conflict nostalgia, a kind of willful amnesia of the violence of the past, partly explains the willingness to turn up the temperature on the language of domination, victory, and absoluteness, and discard the language of generosity, grace, and understanding. As we have long learned, words matter, and those in public life should choose them carefully in a post-conflict society. 

Finally, the apparent ease with which key actors have been careless and defiant of legal process and legal determinations reminds us that once we discard the commitment to our (imperfect) rules, we risk violence, threat, and intimidation as the lingua franca going forward. When political leaders appear to defy the law, the message to all is that the law is irrelevant to our sectarian politics. In post-conflict society, sometimes the imperfect rules are all we have, and to be cavalier with them invites mob rule — or its online equivalent, generated outrage via social media.

So, how to prevent the unravelling of a carefully constructed, if imperfect, peace process now? There is no simple fix. It requires patient back-to-basics, accepting the compromise that was crafted in 1998 and going back to making it work with some humility and grace. There are no shortcuts on a peace process; it is an intergenerational endeavor. It requires accepting that that initial compromise was only the first, and that a successful peace process requires the practice of compromise over and over again. 

The two dominant political parties in Northern Ireland both need tutorials on compromise. Communities need to practice compromise too, embedding ordinary peace in acts of neighborly co-existence that may not be immediately reciprocated. Lasting success still requires two governments, Ireland and the U.K, that agreed to act as peace guarantors, to step up and remain committed, despite the distraction of Middle East wars and an imperious and unpredictable United States that can no longer be relied upon as a neutral guarantor. 

Success also requires making Northern Ireland work, and that means placing a political priority on the things politicians simply don’t want to work on: agreeing on a budget, fixing the health service to better serve the needs of a post-conflict population, finding ways to educate our children together, looking after the elderly who survived conflict and deserve better. 

None of this is sexy. It is definitely less appealing than sitting comfortably in the chair of name-calling and one-upmanship. Most urgently, we also must squarely face the “the cost of our troubles,” because confronting the ugly truth of violence, sectarianism, and grief of the past would remind us why we never want to go back there. This is the hard stuff of building a shared society. But avoiding it, as the Garvaghy Road standoff demonstrates, may simply de-transition us from the future back to the past. 

The post De-transition: Northern Ireland and the Ongoing Challenges of Keeping the Peace appeared first on Just Security.