The plaintiff is Gellert Global Group Corp., a company specializing in imported food products. According to the complaint, "Gellert imports Canadian cheese covered by the Section 338 Tariffs, ... specifically more than 15 product lines of cheeses, including sheep’s milk cheese, mascarpone, borgonzola, chèvre-style goat’s milk cheese, brie, and camembert." In 2025, Gellert imported "roughly $71 million worth of Canadian cheese."

In terms of the big picture here, the complaint argues that "[t]he Section 338 Tariffs are the Administration’s latest attempt to exercise, through a statute never designed for the purpose, the tariff power that the Constitution assigns to Congress." Section 338, it says, is "a relic of the retaliation-based tariff policy that Congress abandoned four years after Smoot-Hawley became law."

As to the specific legal claims, I'll give a quick rundown and offer a few comments below.

Count I: Section 338 has been impliedly repealed or, at a minimum, is controlled by Section 301 of the Trade Act of 1974.

85. Where provisions in two acts are in irreconcilable conflict, the later act, to the extent of the conflict, constitutes an implied repeal of the earlier one; and where a later act covers the whole subject of an earlier one and is clearly intended as a substitute, it operates similarly as a repeal. Posadas v. Nat’l City Bank, 296 U.S. 497, 503 (1936). And a specific policy embodied in a later federal statute controls the construction of an earlier one even absent express amendment. FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 143 (2000).

...

88. Each Canadian practice the Section 338 Proclamations target falls within Section 301. The Dairy Proclamation finds that Canada “denies to the United States the favorable treatment that Canada provides to the EU,” that the practice is “unreasonable,” and that it burdens U.S. commerce. Proclamation 11047 ¶¶ 5–6. And a USMCA panel has found the Canadian rule it targets “not inconsistent with” the USMCA. See supra ¶ 54. The Alcohol and Motor Vehicles Proclamations target measures that allegedly burden or restrict United States commerce. 19 U.S.C. § 2411(b)(1).

89. Section 338 has therefore been impliedly repealed. At a minimum, Section 301 controls where, as here, the President seeks to respond to foreign practices within Section 301’s scope, and the President may not impose duties in response to such practices while disregarding the procedures and limits Section 301 imposes. Because the President did not proceed under Section 301, the Section 338 Tariffs are unlawful whether or not Section 338 survives.

...

When I first heard this kind of claim talked about as a possibility, I was skeptical, but maybe there's something more here than I realized.

Count II: The Section 338 Action exceeds the authority granted by Section 338 of the Tariff Act of 1930.

...

95. None of the Canadian measures identified in the Section 338 Proclamations constitutes discrimination within the meaning of Section 338. Canada’s motor vehicle surtax and its provinces’ alcohol purchasing decisions are responses to tariffs that the United States imposed on Canada, not discrimination by Canada against the United States. And Canada’s dairy TRQ administration treats U.S. cheese better than the cheese of most of the world; that Canada grants the European Union a different preference under a different reciprocal agreement, while giving the United States the TRQ terms the United States itself negotiated and Congress approved, is not discrimination against the United States. Nor is it “unreasonable”: preferences granted under reciprocal trade agreements are a routine feature of trade law, including that of the United States, and a USMCA panel found Canada’s rule “not inconsistent with” the USMCA.

One concern I have with this claim is that while it's true that the Canadian provinces’ alcohol purchasing decisions "are responses to tariffs that the United States imposed," does that necessarily mean they do not constitute discrimination? There's going to have to be a big discussion of what "discrimination" means.

Similarly, yes, the Canadian preferences to EU cheese are part of a reciprocal agreement, but does that mean it's not "discrimination" for the purposes of Section 338? The complaint says: "When Congress enacted Section 338, a preference granted under a reciprocal agreement was not 'discrimination.' The same Act expressly preserved the United States’ own exclusive preference for Cuba." I have no doubt it's true that Congress believed it could still discriminate through tariff preferences, but I'm not sure that meant it was OK with other governments giving such preferences (see the discussion of reciprocity treaties here).

The complaint also raises the issue of whether the Section 338 tariffs properly "offset" the discrimination:

96. Even where its predicates are satisfied, Section 338 authorizes only duties that the President “determine[s] will offset” the burden or disadvantage caused by the foreign country’s discrimination. 19 U.S.C. § 1338(d). Whatever products the President chooses to tax, the duty must balance the burden the discrimination places on United States commerce.

97. The Section 338 Tariffs do not. Each proclamation imposes the statutory maximum rate of 50 percent without tying it to any measured burden: the Alcohol and Motor Vehicles Proclamations recite very different export declines, the Dairy Proclamation quantifies none, and all three impose the same rate. Each states that the duties will “expand opportunities for U.S. producers” and “may spur Canada to remove” the challenged measure, which describes protection and leverage, not an offset. See supra ¶ 56. Proclamations 11064 and 11065 then substituted substantially different product lists while declaring that the duties “would still offset the burden or disadvantage,” and taxed Canadian cheese to “offset” Canadian liquor policies and a Canadian surtax on automobiles. Proclamation 11064 ¶ 5; Proclamation 11065 ¶ 5. They are not offsets; they are general import taxes on selected Canadian goods.

This part of the claim seems pretty strong to me.

The major questions doctrine also comes up here:

98. Defendants’ interpretation of Section 338 also presents a question of vast economic and political significance. Under Defendants’ construction, the President may impose tariffs of up to 50 percent on any products of any country that extends better treatment to any other country in any respect—including through any of the hundreds of preferential trade agreements to which the United States is not a party—or that responds to tariffs the United States has imposed on it. Section 338 contains no clear statement authorizing that power. The major questions doctrine and the canon of constitutional avoidance require Section 338 to be construed not to authorize the sweeping power Defendants claim. See West Virginia v. EPA, 597 U.S. 697, 716 (2022); Learning Res., 607 U.S. at 251.

99. The Section 338 Proclamations therefore rest on a clear misconstruction of the governing statute and constitute action outside delegated authority.

I don't have any great insights into "major questions" issues, or the "nondelegation doctrine" one below. Maybe as the briefs flesh these out I'll come up with stronger views.

Count III: If Section 338 authorizes the challenged tariffs, it violates the nondelegation doctrine.

102. The nondelegation doctrine requires Congress to supply an intelligible principle that meaningfully guides and constrains the exercise of delegated authority. See J.W. Hampton, Jr. & Co. v. United States, 276 U.S. 394, 409 (1928); Whitman v. Am. Trucking Ass’ns, Inc., 531 U.S. 457, 472–76 (2001); FCC v. Consumers’ Rsch., 606 U.S. 656, 673 (2025).

...

103. Section 338 authorizes action only upon a factual finding of discrimination against United States commerce relative to the commerce of other foreign countries and only through duties calibrated to “offset” the resulting burden or disadvantage. 19 U.S.C. § 1338(a), (d).

104. Defendants’ asserted construction of Section 338 eliminates those potential limiting principles. Under that construction, the President may select any products of a foreign country, and tax them at any rate up to 50 percent, without any calculation of the burden to be offset—and may then exclude those products from the United States entirely. And he may do so based on measures that the foreign country adopted in response to tariffs the President himself imposed

Count IV: The collection of Section 338 Tariffs on articles first enumerated in Proclamations 11064 and 11065 violated Section 338’s 30-day waiting period.

110. Section 338 provides that “thirty days after the date of such proclamation there shall be levied, collected, and paid upon the articles enumerated in such proclamation” the new or additional duties. 19 U.S.C. § 1338(d). Section 338 does not authorize the collection of a duty on any article before 30 days have passed from the date of the proclamation that enumerates it.

111. The President observed that requirement in the original Section 338 Proclamations, which issued on July 20, 2026, and were scheduled to take effect on August 19, 2026.

112. Proclamations 11064 and 11065, issued on September 8, 2026, enumerated articles not previously subject to the Section 338 Tariffs, including Canadian cheeses Gellert imports, and directed that duties on those articles be collected beginning September 15, 2026—only seven days later. CBP collected Section 338 Tariffs on those articles, including from Gellert, on entries made on or after September 15, 2026, and before October 8, 2026. Gellert paid more than $381,000 in Section 338 Tariffs on such entries, and it paid no Section 338 Tariffs on its Canadian cheeses before September 15, 2026.

This claim seems like a strong one, but the remedy would likely be just to impose the tariffs in question again after the required waiting period.

Following months of escalating trade tensions, the EU and China reached an agreement on several points of contention during their second meeting on bilateral trade and investment on 8-9 October. One of the key issues at stake concerns imports from China of hybrid vehicles. China’s rise as a global leader in electric vehicle (EV) production and exportation has imposed unprecedented pressure on traditional auto powerhouses including the EU, provoking the adoption of temporary protective measures (e.g. tariffs) and longer-term industrial policy in the pursuit of a systemic overhaul and upgrade of local auto sectors (e.g. EU’s Industrial Action Plan for the European automotive sector). The EU’s countervailing duties, imposed on Chinese EVs in 2024, has led to a phenomenal growth of imports of hybrid vehicles from China while Chinese EV imports flattened. As European car makers announced job cuts in response, this has fueled fears of deindustrialization in the EU and brought up the issue to the top of the EU’s trade agenda since the beginning of this year.

As reported in specialized media, the two sides agreed to limit imports of hybrid vehicles from China into the EU at around 350,000 vehicles per year over the next four years. They indicated that this would be implemented, not through export restraints as initially suggested by the EU, but through safeguard measures. Given that voluntary export restraints are generally prohibited under Article 11 of the WTO Agreement on Safeguards, they claimed that such an approach would ensure WTO-consistency.

Safeguard measures are one of the WTO-permitted trade restrictive instruments for governments to shield domestic industries from import competition temporarily. They are meant to address situations where trade is fair but an increase in imports resulting from unforeseen developments causes or threatens to cause serious injury to a given domestic industry. Where these conditions are fulfilled, a WTO Member can impose a safeguard measure, by way of tariffs, quotas or other forms, to restrict imports irrespective of their sources (in other words, erga omnes). In contrast with anti-dumping and countervailing measures which have strict rules on how to design the measure and its level, safeguard measures can be applied to the extent necessary to prevent or remedy the injury concerned and to facilitate structural adjustment of injured industries, thereby leaving more discretion to the imposing WTO Member. The rules of the WTO Agreement on Safeguards are incorporated in the EU’s Basic Safeguard Regulation.

Since the bilateral meeting, the EU has publicized widely that it will launch a safeguard investigation on imports of hybrid vehicles to impose (provisional) safeguard measures by December this year. EU officials have advertised that the measures would take the form of a quota based on the level of imports into the EU in 2025 plus a mark up of a few percentage points to account for growth in demand. This quota will apply to imports from all countries but the largest exporters to the EU (China, Japan, UK, Korea, US and Türkiye) will get country-specific quotas. Other exporters will share a residual quota.

The EU’s safeguard measure, as proposed, raises several issues. From a legal point of view, it is difficult to see how a safeguard measure, pre-determined through an agreement between the EU and China, can be legal under the WTO Agreement on Safeguards and the EU Basic Safeguard Regulation. Such an ex ante agreement means that the perceived measure will be put in place regardless of whether the legal conditions are met. Given the pre-determined outcome, it is likely that the investigation will be performative only paying lip service to these requirements and the due process required for interested parties. From a diplomatic point of view, the EU’s trading partners, particularly traditional car exporters, were left out of the negotiations and will have little influence on the safeguard investigation. New developments from the investigation that benefit other trading partners may change the dynamics of the EU-China relations and necessitate an adjustment of the negotiated outcomes.

EU officials have claimed that this course of action will benefit trading partners by protecting their market share in the EU. However, since 2025, imports of hybrid vehicles from China have already soared. This is why China agreed to this deal to receive the highest quota and maintain a larger market share than its major competitors. The base rule in Article 5 of the WTO Agreement on Safeguards is that a safeguard measure taking the form of a quota should be established based on the average of imports in the last three years before the investigation. As there were little imports from China in 2023 and 2024 and more imports from other trading partners, this would have not suited China but could have been more agreeable to other car exporting countries (except Türkyie and not however that taking 2026 into account instead of 2023 in a three year average would have further benefitted China):

Furthermore, from a practical point of view, a quota on imports of hybrid vehicles is likely to limit imports at level below that of the quota. This is so because once the relevant quota is exhausted, no more imports can take place. Since the sea voyage can take months, there is no guarantee that sufficient quota quantity would still be available when the vessel arrives at the EU. Given the high cost of car-carrying vessel, car manufacturers will likely act on the side of caution and not start the journey if importation cannot be guaranteed.

Having avoided a full confrontation with China, the EU may nonetheless face one with its other trading partners by purporting to comply with WTO rules in form but not in substance and spirit. The EU already followed a similar path with regard to steel some months ago, leading to tensions. If such transactional methods become the EU’s standard playbook in response to intensified geoeconomic competition, it will set a bad template further undermining the already chaotic global trade order.

Yet, this does not have to be so. The EU could limit imports from China at 350,000 vehicles as agreed with China without affecting its other trading partners.

In the purported safeguard investigation, the EU Commission can apply the principle of parallelism embedded in Article 2.2 of the WTO Agreement on Safeguards (Appellate Body Report, US – Line Pipe, paras. 179-181). This principle provides that a WTO Member can exclude certain WTO Members from the scope of a safeguard measure as long as imports from such WTO Members are not taken into account when assessing whether the criteria to impose the measure are met. The EU has relied on this principle in the past, for example, to exclude the members of the European Economic Area as well as, temporarily, certain countries which had a free trade agreement with the EU mandating exclusion from its safeguard measures on steel in 2018. In the present case, the EU could similarly exclude its free trade agreement and customs union partners from the scope of the analysis and subsequently of the relevant measures. This would allow the EU to keep China in but exclude the UK, Japan, Korea and Türkiye.

One typical challenge here arises from the fact that Chinese car manufacturers have established production bases in third countries such as Türkiye, Morocco and Vietnam. Consequently, targeting solely China and excluding trading partners from the scope of a safeguard measure may render the measure less effective as Chinese car manufacturers can switch from exporting from China to exporting from a third country. Indeed, the EU has many free trade agreements including with the countries just mentioned.  A parallel issue is European car makers themselves having delocalized production outside of the EU. Morocco and Serbia are prime examples of this where French and Italian car makers have invested. Excluding such countries with which the EU has trade agreements may thus simply incentivize European car makers to continue in that direction rather than maintaining production in the EU.

One way to address this issue would be for the EU to draw a line between trading partners which are excluded from the scope and others. This could be done based on the type of trade agreements the EU has with third countries (free trade agreement, association agreement, economic partnership agreement, etc.) or other objective criteria, although this may lead to claims of discrimination.

In any event, the solution proposed above would limit the fall out with trading partners which are not the cause of the issue and were not invited to the negotiating table. It would raise the same issue in terms of paper compliance with the WTO Agreement on Safeguards but would at least limit the number of partners affected.

Another option could be for the EU to adopt a specific regulation to do so through the ordinary legislative procedure (involving the Parliament and the Council) on the basis of the EU’s competence in trade (Article 207 of the Treaty on the Functioning of the European Union) as it did when it raised duties on certain products from Russia and Belarus in the midst of the war against Ukraine. This approach would likely be WTO-inconsistent except if an exception is available, but perhaps now is the time to drop the façade of WTO compliance when such compliance is not achievable in order to maintain the integrity of the system where it can still be preserved.

On 18 September 2026, the European Commission issued Commission Implementing Regulation (EU) 2026/2088 “imposing a definitive anti-dumping duty and definitively collecting the provisional duty imposed on imports of certain alkyl phosphonic acids and their sodium salts originating in the People’s Republic of China”. The anti-dumping duties range from 156.7% to 192.2% ad valorem.

The product concerned (“PBTC”) is produced in only three countries in the world: China, Germany, and India. During the investigation period, PBTC was manufactured by one single producer in Germany (Lanxess), which also acted as the complainant in the investigation (Provisional Regulation, para. 190). Imports from India accounted for only 0.1% of the Union consumption during the investigation period (Provisional Regulation, para. 251). As regards Chinese imports, the Commission found that their volume and total market share increased during the investigation period, while import prices decreased (Provisional Regulation, paras. 197 - 203).

Total Union production of PBTC during the investigation period – that is, the total production of Lanxess – was established at around 7,200 – 9,600 tonnes (Provisional Regulation, para. 190). During the same period, Union consumption was estimated at around 11,300 – 15,000 tonnes, a level therefore much higher than domestic production. Nevertheless, contrary to the arguments advanced by users, the Commission found that the complainant had sufficient spare capacity to absorb a contraction in imports (Provisional Regulation, para. 283). It is also worth mentioning that, during the injury period, not only did the Union industry lose market shares to foreign competitors in the European market, but it also experienced a sharp contraction in its exports to third markets (Provisional Regulation, paras. 252 – 254), showing signs of declining competitiveness.

When examining whether (high) anti-dumping duties were in the Union interest, the Commission dismissed all the arguments advanced by interested parties. Specifically, the Commission concluded that PBTC price increases for European users would not be a problem, as they would “merely be a manifestation of the fair-trade restoration” (Definitive Regulation, para. 80). In addition, the Commission held that, in the absence of measures, the very existence of a manufacturer of the product concerned in the European Union would be endangered (Definitive Regulation, para. 80).

The Commission concluded its analysis with the following enigmatic passage: “[T]he current geopolitical tensions only underscore the importance of local supply sources” (Definitive Regulation, para. 80). This last assertion is not backed by any analysis, nor is it followed by an explanation. Neither the Provisional Regulation nor the Definitive Regulation clarifies whether arguments in this respect were raised by any of the interested parties.

This investigation is illustrative of the Commission’s attitude towards politicizing trade remedy investigations and their outcomes. In an investigation concluded earlier this year, the Commission found that, whereas anti-dumping duties could force users of the product concerned to shut down production facilities or relocate outside of the European Union, anti-dumping measures were still in the interest of the Union, as supporting domestic producers would contribute to preserving the “strategic autonomy” of the European Union. We examined that determination elsewhere, pointing to contradictions and flaws in the Commission’s analysis. Recently, references to “strategic autonomy” were also included in the Provisional Safeguards Regulation on GOES (para. 73), showing that the Commission is willing to rely on these considerations in trade-remedy investigations on imports from all third countries, not just China. In the present case, the Commission merely made an unsubstantiated remark concerning “geopolitical tensions”. As such, we do not know how the Commission considers the product concerned (mainly used in water treatment processes and industrial and institutional cleaning) to be linked to geopolitical interests and, ultimately, to the security interests of the European Union.

In the specific circumstances of this case, the reference to “geopolitical tensions” was an unnecessary additional element in an analysis that already saw the Commission conclude in favour of imposing anti-dumping duties. Nevertheless, the increasingly assertive attitude of the Commission deserves close scrutiny, as geopolitical or strategic autonomy considerations, if not substantiated by a robust, fact-based analysis, may lead the Commission to conclude in favour of remedies even where other elements of the Union interest analysis would suggest otherwise.

Geopolitical tensions are real and will not dissipate any time soon. In view of their importance, they should be handled with care. If the Commission considers that geopolitical considerations should inform its trade remedy decisions, it should spell out the parameters of this analysis ex ante and in clear terms, so as to ensure due process and promote transparency in its decision-making processes. When geopolitical considerations are not necessary, either because there is already enough evidence on record or because there are no elements to suggest that the product under investigation is strategic, the Commission should exercise restraint and limit itself to the core features of trade remedy investigations.

*International Trade Lawyer (ccantore@pm.me).

Recently, this has come up a lot in the context of tech regulation, with accusations of protectionism and responses indicating that there is a legitimate public policy being pursued. Here's a recent example from Australia that is a good illustration of how the issue arises in the real world, as Australia is considering some online safety legislation and the Trump administration is raising objections.

The proposed Australian online safety legislation and the Trump administration's response

The draft legislation in question is Online Safety Amendment (Digital Duty of Care) Bill 2026, which the Australians are in the midst of debating. The explanation of the bill by the Australian Prime Minister's office notes the following elements related to algorithmic feeds and harmful features or content:

• Users will be able to "choose to opt in to having their default feed include personalised content recommended by the algorithm"; "[a]lternatively, users can opt out of personally recommended content in their default feed and see the friends and creators they choose to follow."
• "digital services like online games, apps and AI chatbots will be required to protect under 18s from design features that have negative behavioural impacts, like addictive features or those that can have an effect on a young person’s self-esteem."
• These same services will also be required "to protect under 18s from harmful content like": "content that promotes, encourages or accelerates eating disorders"; "content that promotes hostile ideas about women and gender equality"; "pornography"; "content that glorifies crime or life endangering stunts"; and "content that causes serious mental health distress, including abuse and bullying."

The Australian government asked for feedback on the draft law, with the legislation to be introduced in Parliament this year.

The Trump administration submitted a response that puts forward a wide range of concerns. Here are some of the key arguments it made:

• "We are particularly concerned by potential mandates that would allow regulators to impose rigid, one-size-fits-all platform design requirements – such as forced chronological feeds or opt-in recommendation systems."
• "A framework that empowers governments to enforce vague definitions of 'harm' risks becoming a mechanism for viewpoint-based censorship, as platforms will be incentivized to over-moderate speech that should be protected to avoid regulatory penalties."
• "If Australia compels platforms to aggressively filter content in algorithmic feeds to satisfy a government mandate, and such filtering draws in protected speech, the effect on a disfavored speaker or media source’s reach would not be meaningfully different than outright removal.  The U.S. government would likely consider such a measure as facilitating censorship."
• "Mandated platform design features, especially when applied to algorithms, may affect what users see, say, and hear not just in or from Australia, but globally.  Platforms generally build and deploy product features at a global scale rather than market-by-market; forced platform design changes to mitigate Australia-specific risks could result in systemic changes that affect how users everywhere, including Americans, experience these platforms and reach audiences."
• "The U.S. government has made clear its opposition to foreign measures that result in extraterritorial censorship of protected speech by Americans online."
• "Regulations that impose disproportionate compliance burdens on American companies relative to competitors, or that apply unevenly across similar services, will invariably raise concerns."

I have some concerns with what each side is doing here. I'll start with the Australian legislation and then turn to the U.S. response.

How to improve social media feeds: Regulation of existing companies vs. encouraging new platforms with different business models

Starting with the merits of the draft Australian legislation, I'll focus first on the "potential mandates" for "forced chronological feeds or opt-in recommendation systems." I agree that chronological feeds and an opt-in for recommendation systems are good features for social media, and I'm happy that Australia is trying to promote them. I've never liked having someone else decide what I see on my social media feed, especially when this is done in a non-transparent way, as is generally the case. (I soured on Facebook pretty early when I realized they weren't showing me every post from people I followed, but rather using an algorithm to decide what I saw.) As a I result, I agree with the motivation behind the Australian draft as it relates to this issue.

However, I'm skeptical that government regulation is the right approach to addressing the problem. My concern is that regulation along the lines of what Australia is proposing will leave us with the same companies offering pretty much the same problematic services, with minor tweaks that don't get at the issue of the standard social media business model that is at the heart of the problem: Maximize usage of the platform, gather data on users, and sell that data to advertisers.

In my view, a better model for social media companies would be to offer chronological feeds, either as the exclusive option or as one of multiple options, and look for alternative sources of revenue besides the collection and sale of data. For me, the most useful thing the Australian government and other governments could do to move things in this direction would be the following:

• Use the bully pulpit to encourage companies to offer chronological feeds or feeds that people can customize in a transparent way.
• Use the bully pulpit to encourage their citizens to use chronological feeds or customized feeds.
• In their own social media activities, use services that take this approach rather than services that rely on non-transparent company-provided algorithms. And if they find the existing range of social media services inadequate, encourage companies to build new ones around a different model.

The goal here is to shift the market in the direction of new business models. A model based on toxic engagement and selling user data is one way to make money in this market, but it is certainly not the only one, and it would be nice if we had more options.

I'll go even further here and say that legislation of the sort Australia is proposing could distract from what is needed. When politicians make claims that regulation will address a problem, it causes people to think the problem is being solved for them, and steers them away from doing what needs to be done themselves: Building and using alternative platforms, ideally ones that are interoperable with each other (i.e. decentralized social media) rather than the "walled garden" platforms many people use today.

Finally, as to governments trying to influence what shows up in algorithmic feeds, which is also part of the Australian draft legislation, I think that while it is well-intended, it may not work out as well as people hope. Each change in government could mean a shift in what exactly government officials are worried about people seeing in their feeds, with these feeds turning into an endless back and forth between right-wing and left-wing agendas. In my view, regulating the content of company-provided algorithms is unlikely to go well, and I think this supports my earlier point that we should all try to avoid this type of feed.

Pushing back on U.S. government pushback against foreign tech regulation

While I'm not thrilled about aspects of the Australian legislation – because I see what they are going for but I question whether it is the right approach to getting there – I'm even less enthusiastic about the Trump administration's response. As far as I can tell, none of what Australia is doing is intended to discriminate against American products companies. As a result, I'm not convinced the U.S. government should be pushing the issue with the Australian government in the way it is doing here.

In my view, governments should tread carefully when offering criticism of foreign regulations. All regulations can affect the interests of exporters from the contesting government's country, but the impact of these regulations varies considerably, ranging from an equal impact on all companies, a disparate impact on foreign companies, and a disguised attempt to protect domestic producers from foreign competition. Here, there is nothing to suggest that the Australian legislation should be thought of as disguised protectionism, and I think that is an important consideration.

A key part of my thinking here is a concern with whether international trade principles intrude too much into domestic policy-making. If we rely on a principle of international trade under which "disproportionate compliance burdens" are sufficient to constitute an actionable trade barrier, it could lead to trade conflicts that cannot easily be resolved. When a government is regulating in a way that it genuinely sees as in the public interest, objections by other governments come across as an intrusion into sovereignty or policy space. As we've seen over the years, that often goes badly.

Obviously, the affected companies themselves can express their views to the Australian government, and they should probably do so here. If you are operating in the Australian market, you have an interest in these matters, and can offer your perspective. But other governments doing the pushing in the form of complaints about "market access" and "trade treatment" is something very different, and has broader implications for the overall functioning of the trading system (the administration's response says "the proposed digital duty of care regime, depending on how it is implemented, potentially raises significant concerns regarding fair competition, market access, and trade treatment for American companies, independent media, and content creators").

Turning to the "censorship" issue, I raised some concerns above with the Australian approach (although I'm not sure I have a good sense of what this approach would mean in practice), but I have even stronger concerns with the Trump administration's reaction.

One issue is that there's plenty of censorship going on in the U.S. right now, so making allegations about Australia's censorship feels odd.

Another issue is that speech policies are always going to vary by country, based largely on the views of the people of that country as to what speech should be allowed. Americans may prefer a certain approach at the moment (or maybe they don't even like the current approach, as the policies are set by the Supreme Court to a great extent), but that's not an argument for telling other countries how they should do things on speech policy, and I think the U.S. government should be very careful about telling others what to do.

That's not to say it should do nothing about concerns related to free speech in other countries though. What I think would be useful here is if there were an international body that served as a forum for governments to talk about comparative speech policies and the extraterritorial effect of domestic speech policies. Everyone would probably learn a lot from that exercise.

In terms of the impact of the draft legislation on U.S. companies, I can understand how this sort of regulation is a big pain for companies who operate in different markets, and who will have to tailor their services for local statutes and regulations. But that's just how things work, and they will have to figure out how to deal with it. If a particular company wants to follow only American laws, it can do so: Just offer the service only in the U.S. The U.S. is a big market, so there's still plenty of money to be made if they don't want to adapt to foreign regulations. On this point, the Trump administration says:

Platforms generally build and deploy product features at a global scale rather than market-by-market; forced platform design changes to mitigate Australia-specific risks could result in systemic changes that affect how users everywhere, including Americans, experience these platforms and reach audiences.

That may be true where you have walled garden platforms operating globally, but that's not a great system and this approach is in need of disruption. What would be better is if we had decentralized social media, in which all of the platforms could connect, and you had U.S. companies following U.S. law, Australian companies following Australian law, etc., with users in each place still being able to connect with each other.

Finally, the Trump administration is also concerned about Australia forcing companies to do filtering of algorithmic feeds. To me, this is just more evidence that algorithmic feeds have some fundamental flaws, and people should consider alternatives to company-provided feeds (and companies themselves should deemphasize them). I know there are some people who prefer algorithmic feeds to chronological ones, although despite many conversations I've had on this issue, it's still not clear to me why they prefer this. Regardless, I do recognize that there is some demand for these feeds, and what I would suggest here is that if companies are going to offer them, they should give users the ability to configure them so that people have a better sense of what they are getting.

Conclusion

Summing all this up, what I would like to see here is Australia take its concern about the impact of social media feeds and online content, which I think is a legitimate worry, and channel it into something a bit more productive. As the same time, I wish the U.S. would focus on rooting out disguised protectionism in foreign tech regulations rather than complaining about any and all regulations that have an impact on U.S. companies or reflect different values than those of the administration in power at that moment.

UPDATE:

This issue came up in a recent interview with Australian Trade Minister Don Farrell:

Host: Digital duty of care laws are going to be introduced this week. Back In September, the U.S. embassy, you'll know this, made a rare intervention into our political debate, calling the laws extraterritorial censorship of protected speech. Have you spoken about those laws or have they been raised with you, with your counterpart?

Minister: Yes, they've been raised with me every time I've met with the Americans. We obviously have a difference of opinion. We certainly don't believe if anything we are doing is contrary to our Free Trade Agreement. But it's the job of the Australian Government to protect Australian citizens. There's a lot of terrible stuff out there in social media. I've tried to explain to the Americans that our objective is to protect our citizens, particularly young people, and so far there hasn't been a direct response.

Host: Has there not been a threat of a consequence for these laws?

Minister: No, no, they have never raised.

Host: They've raised it with you. Is this Jamieson Greer?

Minister: Yes, Jamieson Greer.

Host: And he's actually raised the digital duty of care directly with you.

Minister: He's raised those issues, a range of issues in this social space. And each time I explain why we're doing this, and he certainly hasn't raised any issue. In a few minutes, I'm meeting the new American ambassador.

Host: You suspect it will be raised again?

Minister: Oh, he may or he may not, I don't know. But if he does, I will explain to him. As I've explained to Jamieson Greer, our job as an Australian Government is to protect our citizens. And this is one of the ways we're doing it.

Congrats to Canada on winning this unofficial competition for fairest trading partner! Canada also scored the highest on the question "Please describe the United States’ relationship with the following countries," with 72% saying it was a "friend and ally." Thus, despite what's going on at the highest political levels, most Americans still appreciate Canada.

(For comparison purposes on the fair trade question, I would be curious to hear the responses in a poll of Americans to the question "Would you say the United States is a fair or unfair trading partner?" Maybe the folks at CFR can add that one in next time.)

Another interesting question was "What do you think the next president should do about Trump's tariffs on the following trade partners?" Here are the results:

This is a complex issue, and I'm not sure how to phrase it to get a good sense of public opinion on where to go with U.S. trade policy under a new administration. Asking only about removal of the tariffs leaves some uncertainty about what else would go along with that. Rather than a country-specific question focusing on tariff removal, I might try something along the lines of the following instead: "Should the next U.S. president remove all of Trump's tariffs and then develop a new overall tariff and trade policy?"

But it's tough to capture everything that is going on with U.S. trade policy at the moment in a poll question. Among other things, you have the tariffs themselves, the various trade deals negotiated by the Trump administration, the choice of countries being targeted, the choice of sectors being protected of affected, and the specific accusations and concerns about unfair trade practices. All of it could be reconsidered by a future administration. I'm not sure polling can get at every aspect, though, and I think that to a great extent a future administration is going to have use their instincts as to the details of what Americans want to see here.