The post Can Tisza restore Hungarian democracy? first appeared on LSE European Politics.

Eleonora Celoria, Post-doctoral researcher, FIERI

Andreina De Leo*, Post-doctoral researcher, Maastricht University

Marcella Ferri, Research Fellow, Florence University

* Funded by the European Union (ERC, SoftEn project, 101165167, PI: Lilian Tsourdi). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them.

Photo credit: Jorge Franganillo, via Wikimedia Commons

Introduction

On 23 and 24 March 2026, the Court of Justice of the European Union (CJEU) held two hearings in the ongoing proceedings addressing the compatibility of the Italy-Albania Protocol with EU law and the broader implications of conducting asylum and return procedures extraterritorially (Joined cases C-706/25 Comeri and C-707/25 Sidilli, and C-414/25 Sedrata).

This may appear as a déjà-vu. Less than a year ago, the extraterritorial processing of asylum under the Italy-Albania Protocol had already come before the Court in the well-known Alace and Canpelli cases (discussed on this blog by Zamboni and Favilli & Marin). The focus of those cases was, however, narrower, as it concerned the criteria to designate countries as “safe countries of origin” under the Asylum Procedures Directive (APD) and their judicial review, a classification that directly enabled the use of accelerated offshore asylum procedures in Albania. Following those rulings, the Italian Government adjusted its strategy, implementing a policy based on the relocation to Albania of returnees already detained in Italian pre-removal detention centers under the Return Directive (RD). While awaiting removal, these individuals may lodge an asylum application which, according to the Government, can then be examined in Albania. Against this backdrop, the new preliminary references raise a distinct and more direct question: whether asylum and return procedures can, as such, be carried out in a third country without breaching EU law. In doing so, they place the legality of the Protocol as a whole squarely before the Court.

More specifically, the first preliminary ruling, submitted by the Court of Cassation on 20 June 2025, asks whether Articles 3, 6, 8, 15, and 16 of the Return Directive preclude transferring a migrant subject to a return order to an extraterritorial detention centre, even without a concrete prospect of removal. It also asks whether Article 9(1) of the Asylum Procedures Directive, which requires applicants to remain on a Member State’s territory, precludes keeping asylum seekers who lodge an asylum application after being transferred to a third-country detention centre within that country (see De Leo).

The second preliminary ruling, submitted by the Rome Court of Appeal on 5 November 2025, directly questions Italy’s competence to conclude an international agreement  interfering with an area of exclusive EU competence, as laid down in Articles 4(3) TEU, 3(2) TFEU and 216(1) TFEU (see Montaldo and delli Carri). Alternatively, should the Court find that Italy has such competence, it asks whether the safeguards under the APD and the Reception Conditions Directive (RCD), as well as the Charter, particularly the rights to liberty, an effective remedy, and health, preclude transferring asylum seekers to, and detaining them in, centres outside EU territory.

It is worth recalling that the stated aim of the Protocol is to “reinforce bilateral cooperation in managing migration flows,” allowing Italy to transfer migrants to Albania solely to carry out asylum and return procedures under Italian and EU law (Art. 4(3) Protocol). These procedures remain under Italian jurisdiction and are conducted in accordance with national and EU law “to the extent compatible” (Art. 4(1), Ratification Law 14/2024).

Although raising different questions, both Italian courts focused on the same core issue: whether EU law safeguards can be effectively ensured for asylum seekers and returnees detained in a third country and whether those standards are compromised merely because detention occurs outside the territory of a Member State.

Outline of the post

While awaiting further developments, and notably the Advocate General’s opinions on 23 April (Sedrata) and 11 June (Comeri and Sidilli), this post examines the key issues raised during the hearings and outlines the main arguments on whether asylum and return procedures conducted in a third country, yet under a Member State’s jurisdiction, comply with EU law.

It first considers whether EU law applies to procedures carried out in Albania and whether such procedures may interfere with its application, summarising the positions of the Italian Government and the European Commission. Two aspects are examined: the Dublin system and its founding principle of mutual trust, and whether EU safeguards can be effectively guaranteed in detention centres abroad. Finally, the post clarifies the differences between the Albania model, the so-called “return hubs,” and the notion of a safe third country, as frequently referenced during hearings and in public debate.

EU Law Beyond Territory: Parties’ Arguments on Applicability, Compliance, and Competence in the Italy-Albania Scheme

The Italian government

To justify the compatibility of the Italy-Albania Protocol and its Ratification Law with EU law, the Italian Government advanced two alternative lines of argument. Both were aimed at demonstrating that: first, the Protocol does not risk adversely affecting the internal rules of the Union and therefore does not encroach upon an area of exclusive EU competence; secondly, that the extraterritorial application of the relevant directives in a third country does not undermine their effet utile.

First, the Government argued that the Albanian centres should be considered “representations of Member States” abroad under the APD and RCD, and that EU law applies indirectly through its incorporation via the Ratification Law. On this basis, the Government maintained that the Italy-Albania Protocol is purely technical and logistical: it does not alter the substance of the legal framework on pre-removal detention, asylum procedures or reception conditions, which remain fully applicable because the legislator has chosen to extend EU standards to the situations covered by the Protocol. By virtue of this legal extension, the Government maintained that third-country nationals detained in Albania receive treatment equivalent to that in Italy. It further argued that instruments such as the Dublin Regulation, though not explicitly mentioned in the Protocol, continue to apply because the individuals remain continuously under Italian jurisdiction. Accordingly, the Protocol would neither interfere with the Common European Asylum System (CEAS) nor encroach upon the Union’s external competences.

With respect to the RD, the Government argued that the transfer to Albania does not qualify as a “return” within the meaning of Article 3, but it rather amounts to a temporary relocation to a different pre-removal detention facility, legally assimilated to ones located on Italian territory. Accordingly, it maintained that compliance with the Directive’s objectives and provisions, including Articles 15 and 16 on detention and conditions of detention, are not undermined, as the material conditions would be fully respected. Regarding the APD, the Government contended that Article 9 is not violated when returnees apply for asylum after being transferred to Albania, since they remain under Italian jurisdiction, and thus do not risk refoulement, and continue to benefit from the full application of both the APD and RCD.

Alternatively, the Government argued that EU law could apply directly through a functional interpretation of “territory.” In the absence of a clear EU definition of “border” or “transit zones,” the centres in Albania could be regarded as a fictio iuris, functionally equivalent to transit zones located on Italian soil. On this view, the legal equivalence of procedures in Albania and Italy would justify treating the Albanian centres as falling within the scope of EU law, in the same way as their territorial counterparts.

While the Dublin Regulation was not explicitly mentioned by the referring judges, it emerged as a contentious issue. In particular, the risk of circumventing Dublin criteria for unaccompanied minors (Article 8) and family members (Articles 9–10) was debated during the Sedrata hearing. The Italian Government maintained that the Regulation remains applicable because the Albanian centres qualify as “transit zones” under Dublin III. This, however, raised a fundamental question from the bench: on what legal basis could another Member State be required to accept a Dublin transfer from a non-EU centre, and how would this not interfere with common EU rules? Questions which were left unresolved.

The European Commission

The Commission, by contrast, adopted a strict literal reading of the directives, distinguishing clearly between jurisdiction and territory. While acknowledging that the individuals remain under Italian jurisdiction, it stressed that asylum and reception rules are territorially anchored and do not apply automatically outside Member State territory. Nonetheless, the Commission did not view the Protocol as violating EU law or encroaching on exclusive EU competences, so long as the legislator’s choice to apply EU standards effectively preserves the effet utile of the acquis. Its reasoning followed two separate lines, depending on the instrument in question.

As for the RD, the Commission noted that it does not strictly define its scope in territorial terms but rather sets standards to ensure the effectiveness of returns from EU territory. Since Italy has committed to applying these standards in Albania, the Commission did not see any inherent incompatibility with the temporary transfer of returnees outside the territory pending repatriation. Yet, this conclusion was subject to strict conditions: the Commission emphasised that detention should have been previously judicially authorised, that all guarantees under Articles 15–18 of the Directive are fully respected, and that the transfer to Albania must not constitute the execution of the return decision. In essence, for the Commission the Directive does not require detention to occur on the territory of the Member State: what is rather decisive is that the Member State fully complies with its concrete obligations.

As for the APD and the RCD, the Commission rejected the qualification of the Albanian centres as “transit zones.” It stressed that the notion of territory is autonomous under EU law and cannot be extended to third countries, and that thus a “transit zone” must be located within the Union’s territory. When pressed by the bench on whether an international definition of “transit zone” exists, it indicated that this was ultimately irrelevant, as the notion must be understood as a matter of EU law. Accordingly, the Protocol cannot extend EU territory, and the direct applicability of the asylum acquis is thus excluded.

That said, the Commission did not view the mere extension, via national law, of EU standards to procedures in Albania as automatically undermining the directives’ objectives or interfering with internal EU competences under the ERTA doctrine. Such a violation would only arise if the arrangement risked affecting common EU rules: a risk the Commission considered unproven here. Article 9 APD illustrates this approach: if the asylum application is made in Italy, transfer to Albania is precluded. Conversely, if lodged only after transfer, while the Directive does not apply directly, its purpose, i.e. to prevent removal before assessing refoulement risks, is fulfilled because Italy retains jurisdiction and applies EU standards.

Regarding the Dublin system, the Commission held that the Regulation does not apply outside EU territory, revealing a potential paradox noted by Advocate General Emiliou during Sedrata: if EU law does not formally apply, can it still be circumvented?

More broadly, the question arises whether exclusion from the scope of the Regulation and Directives is merely a consequence of its incidental inapplicability in a third country, or the result of a deliberate choice by Italy, and namely, transferring individuals to Albania. In the latter scenario, how can it be argued that the Protocol and Ratification Law do not effectively allow Italy to evade its obligations under EU law?

As President Lenaerts highlighted, the key concern of the ERTA doctrine is preventing Member States from using external bilateral agreements to regulate matters that fall within EU law, in ways that could undermine its objectives. In other words, it is not enough to claim that EU law does not directly apply in a third country, but what matters is whether the external action could affect the effective internal functioning and uniformity of the EU legal framework. In this respect, the Commission’s position appeared unsatisfactory, as it failed to address the broader systemic implications for the CEAS and the structural risks of circumvention inherent in the scheme.

This brings us to our legal assessment, where we examine the unresolved issues in the approaches of the Italian Government and the Commission regarding the challenges of extraterritorialising asylum procedures and applying EU law in a third country.

Assessing the Potential Interference of Procedures Conducted in a third country with the EU Migration and Asylum Acquis

All in all, both the Italian Government and the European Commission ultimately arrive at a broadly similar conclusion: the Italy-Albania Protocol does not entail a violation of the Union’s exclusive competences, nor does it compromise the effet utile of the relevant directives, insofar as EU standards can be ensured, even in a third country. However, this conclusion leaves several key issues unresolved. First, it does not fully address the implications for the Dublin system. Indeed, the extension of Dublin obligations outside the Union could highly interfere with the CEAS and the principle of mutual trust. Second, the assumption that equivalent standards can be guaranteed extraterritorially remains highly problematic, as the territorial scope of EU asylum law reflects structural constraints necessary to ensure full compliance with procedural and substantive guarantees. These two issues will be addressed separately in the following sections.

The (application of the) Dublin system: a paradigmatic interference with the CEAS

The application of the Dublin system emerged as a highly sensitive issue during the Sedrata hearing. While EURODAC was not discussed, the reasoning for the Dublin Regulation can be extended to it, in light of recital 54 of EURODAC Regulation. The Italian Government and the Commission took sharply divergent positions: the Government qualifies the centres as “transit zones” under Dublin, a view opposed by the Commission, which raises several critical questions.

First, how can the Albanian centres be considered “representations of Member States” under the APD and the RCD, or “transit zones” under the Dublin Regulation? This argument is unpersuasive, as it would imply that the notion of “transit zone” differs across two CEAS instruments, risking to undermine the system’s internal coherence.

Secondly, if the Dublin system were applied to asylum requests lodged in the Albanian centers, other Member States would be obliged to accept transfers, as the Italian Government argued. From a theoretical perspective, this would directly interfere with the CEAS and undermine the principle of mutual trust on which it rests (Lenaerts, 2017). In fact, the presumption of compliance with EU law standards appears questionable in itself when it is applied to an asylum request lodged outside the territory of a Member State — albeit under its jurisdiction. Yet, save for specific exceptions, that principle prevents other States from verifying whether, in a given case, Italy has actually complied with EU law in processing an asylum application, notably that it has allocated responsibility in accordance with the procedural guarantees laid down in the Dublin Regulation and with the right to an effective remedy against the transfer decision. In short, extending the Regulation to applications lodged in the Italian centres in Albania would run counter to the principle of mutual trust underpinning the Dublin system and, ultimately, the CEAS.

Third, from a practical perspective, applying the Dublin Regulation outside Italy would adversely affect the rights of asylum seekers lodging an application in another Member State and seeking reunification with a family member detained in Albania (Article 10 of the Dublin Regulation). Unless the applicants were also transferred to Albania — which would amount to their de facto detention — they would remain in Italy, thereby rendering effective reunification  impossible. This impossibility, which stems directly from the detention of the family member in Albania rather than in a detention facility on Italian territory, would breach their rights to family life and to the child’s best interests, as laid down in the Dublin Regulation and in Articles 7 and 24 of the Charter.

By contrast, if, as the Commission argues, the Dublin Regulation does not apply, Italy could evade its obligations under Articles 8— 10 of the Regulation and Articles 7 and 24 of the Charter. Once again, such circumvention would be achievable simply through the deliberate extraterritorialisation of asylum procedures. In a nutshell, whether applied or not, the Dublin Regulation illustrates the systemic and structural interference with the CEAS created by the Protocol and Ratification Law.

Ensuring Effective Application of  EU Standards and Rights in a Third Country: Mission Impossibile?

We concur with the Commission that the territorial nature of EU asylum law imposes inherent limits on extending its directives beyond the Union. However, we diverge from the view that voluntarily applying EU standards abroad via national law automatically resolves compatibility concerns. This divergence is anchored in the structural reality that the territorial scope of the directives is not merely formal: it reflects the impossibility of fully guaranteeing procedural and material standards outside the Union. Applying EU law where these guarantees cannot be secured risks undermining the directives’ objectives, the uniformity of EU law, and may raise concerns under the ERTA doctrine (see: De Leo & Celoria, and Montaldo).

As highlighted by the referring judges and the lawyers representing the migrants, a central issue is the gap between the guarantees afforded to migrants detained in Italy and those in the Albanian centres. The Government asserts that procedures in Albania are identical to those in Italy, yet legal equivalence on paper does not ensure effective protection in practice. In our view, the fiction that Albanian centres are legally treated as part of a Member State is insufficient to secure genuine compliance with EU standards on the ground.

The main concerns identified by the referring courts and discussed during the hearings include: the right to liberty under Article 6 of the Charter, notably the requirement for immediate release if detention is unlawful; guarantees related to detention conditions, including access for lawyers, family members, national, international, and non-governmental organisations, and access to healthcare; and the right to a fair trial and effective remedy under Article 47, particularly the right to be “advised, defended, and represented.” As emphasized by the rapporteur judge of Comeri and Sidilli case, these discrepancies risk creating a distinction between “two types of asylum seekers,” or even “two types of detainees,” solely based on where and when an application is lodged.

Both the RD and the RCD require that a person “shall be released immediately” if detention is found unlawful (Article 15(1) RD; Article 9(3) RCD). This requirement is inherently impossible to reconcile with the Italy–Albania scheme. Under Article 6(2) of the Protocol, Italian authorities must “take the necessary measures to ensure the permanence of migrants within the areas, preventing their unauthorised exit into the territory of the Republic of Albania, both during and after the completion of administrative procedures, irrespective of the final outcome.” In practice, release on Albanian territory is impossible: individuals can only be freed once transferred back to Italy. Therefore, even when detention is considered unlawful, the person remains under continuous coercive control throughout the transfer, i.e. from the detention center to the port, during maritime transport, and until release in Italy (typically in Bari or Brindisi): a process that may last from 24 hours to several days, depending on logistical conditions. In this respect, the Italian National Guarantor has recently noted that, if transfer cannot occur the same day, the person remains in a “separate area” of the same detention centre, a practice arguably failing short of the “immediate release” requirement. During this period, individuals are subject to a “coercive measure that deprives [them] of [their] freedom of movement and isolates [them] from the rest of the population” (CJEU, FMS, para. 223). Both the Albanian centres and the transport arrangements display the defining features of detention: a closed, restricted space where movements are limited and monitored and exit is not voluntary (CJEU, FMS, para. 231). The European Court of Human Rights similarly confirms that confinement on ships may constitute de facto detention, even if nominally for the person’s interest (Khlaifia and Others v Italy, paras. 70–71). In sum, the extraterritorial setting alters the legal consequences of detention, making immediate release practically unfeasible and raising serious questions about compliance with EU law.

Extraterritoriality also affects effective access to detention centres. Under national law implementing EU standards (Articles 16(2) and (4) RD; Article 10(4) RCD), a wide range of actors, e.g. parliamentarians, national monitoring bodies, UNHCR representatives, lawyers, family members, religious ministers, and civil society organisations, should be granted access (Articles 21 DPR 394/99; 7 D.Lgs. 142/2015; Ministerial Directive 19 May 2022). By contrast, Article 9(2) of the Protocol limits access to “lawyers, their assistants” and “international organisations and EU agencies,” raising questions about the interplay of the two regimes. Even if Italy seeks to apply its broader framework, as argued during the hearing, effective implementation depends on Albanian authorities’ consent. This is because access requires entry into Albanian territory, which remains under Albanian sovereignty. Article 9(2) subjects access to “applicable EU, national and Albanian law,” while Article 6(1) confirms Albanian responsibility for public order and security outside the centres and during transfers. Consequently, Albania may lawfully restrict access based on domestic entry, residence, or public order rules, limiting in practice the actors able to monitor detention conditions. While the Italian Government argued that this poses no practical concern, citing visa-free entry for holders of Italian passports or residence permits, the effectiveness of EU law guarantees cannot depend on conditions outside the Member State’s control. Making access contingent on the rules or discretion of a third country introduces uncertainty incompatible with the requirement to ensure EU rights in an effective and uniform manner. Moreover, denials imposed by Albanian authorities cannot be effectively challenged before an Italian court, which undermines the right to an effective remedy for those having the right to access the detention centres.

Similarly, under Article 10(4) of RCD, lawyers, family members, and civil society organisations have the right to communicate with and visit detainees, and any limitations must not make access impossible or excessively difficult. In the Albanian centres, however, restrictions appear structural and generalised rather than exceptional, making them difficult to reconcile with Articles 7 and 47 of the Charter and the proportionality requirement under Article 52(1). The Government’s argument that similar limitations exist in Italy is unpersuasive: in Albania, restrictions are the norm, access is more complex, time-consuming, and costly, and a recent monitoring report by the Italian National Guarantor for the Rights of Persons Deprived of Liberty notes that some detainees were not informed of their transfer, further hindering visits. Crucially, as noted above, the ultimate decision rests with Albanian authorities, meaning that limitations stemming from a third State’s sovereignty cannot be effectively challenged or remedied in Italy, thereby weakening the effective enjoyment of EU-law-derived rights.

Additionally, discrepancies in safeguards for lawyers are particularly pronounced. In Italy, appointed lawyers can access detention centres without prior authorisation and often be appointed in person after initial telephone contact. In Albania, this is not possible, nor foreseen by the Protocol or the Ratification Law, which allows travel at public expense only “when remote connection is not feasible” during the detention validation hearing (Article 4(5), Law 14/2024). This effectively precludes in-person meetings before or after the hearing to prepare a defence or appeal a negative decision. Given the current five-day deadline to challenge a validation decision, detainees without immediate lawyer access face heightened risk of ineffective representation. In-person meetings are often essential to identify vulnerabilities or health conditions that may render detention unlawful, tasks difficult to perform remotely, particularly when access depends on the private centre manager’s discretion and mobile phone use is restricted. Physical distance, limited contacts, and communication restrictions materially weaken lawyer-client interaction, undermining the right to effective assistance and representation under Article 47 of the Charter. Remote hearings further exacerbate these issues. By analogy with the Court’s reasoning in FP and Others (C‑760/22), videoconference participation is not inherently incompatible with a fair and public hearing, but its appropriateness depends on the individual being able to follow proceedings, be heard without technical obstacles, and communicate effectively and confidentially with their lawyer. In Italy, lawyers attending remote validation hearings can choose to be in the courtroom with the judge or at the centre with their client: a flexibility that is simply unavailable for detainees in Albania, further undermining effective representation.

Finally, significant disparities exist regarding the right to health, guaranteed under Article 16(3) RD and Article 17(2) RCD. Detainees in Albania face substantial limitations in medical services, particularly psychiatric and addiction care. Unlike in Italy, where detainees are integrated into the National Health System, no equivalent framework operates in Albania, and on-site medical teams cannot provide comparable treatment. These deficiencies have already led to serious consequences, including self-harm and suicide attempts (see, report, here). A recent report by the National Guarantor for the Rights of Persons Deprived of their Liberty found that, of 26 detainees for whom updated medical assessments were requested, 25 were deemed unfit for detention and released, highlighting the centres’ inadequacy to ensure proper care. While Article 4(8) of the Protocol obliges Albanian authorities to provide treatment in emergencies beyond Italian capacity on site, healthcare in a third country is by definition different from that provided by a Member State’s National Health System.

Taken together, the disparities analysed above make detention in centres located in a third country significantly more burdensome than detention in Italy, casting doubts as for its compliance with the principle of proportionality. There is no justification for imposing a harsher coercive measure than would be applied in Italy, particularly as the transfer to Albania does not demonstrably enhance the effectiveness of return procedures under the RD. On the contrary, a report by the Italian National Guarantor for the Rights of Persons Deprived of their Liberty shows that only 56 of 192 individuals transferred to Albania were ultimately returned to their country of origin, representing a return rate of approximately 29%, which is lower than the general return rate.

Conclusion

In conclusion, while the Italian Government frames the issue in terms of formal equivalence between procedures carried out in Albania and those in Italy, and the European Commission emphasizes that the voluntary application of EU standards suffices to ensure compliance, our assessment shows that this approach does not adequately address the structural limitations inherent in the de facto extraterritorial application of EU migration and asylum law.

While awaiting the Court’s decision, we maintain that the strict territorial scope of EU law in this area is not merely formal but reflects a substantive requirement: it is meant to operate within the territory of Member States, where its guarantees can be fully implemented and monitored. Remaining under the jurisdiction of a Member State alone is insufficient to ensure compliance, as EU law cannot be effectively applied outside the Union without undermining its purpose, coherence, and the procedural and material protections that are inherently tied to the territorial context in which they are guaranteed.

Two key issues arise. First, if the Dublin Regulation applies to applications lodged in Albania, the Protocol and Ratification Law would undermine mutual trust and the functioning of the Dublin system, since other Member States cannot verify compliance with EU law outside the Union. Even if it does not apply, the deliberate delocalization of asylum procedures to a third country allows circumvention of EU obligations. Second, procedural and material safeguards ensured in Italy cannot be fully replicated in a third country.

Thus, in our view, relocating individuals to Albania undermines the effet utile of the CEAS, creating tangible discrepancies in treatment that formal equivalence cannot remedy. This, in turn, risks breaching the principle of loyal cooperation and jeopardizes both the uniform application of EU law and the trust-based cooperation that underpins the EU legal order.

A Final Note: Why Albania Is Different from Other Externalisation Models

Throughout the hearings, “return hubs” were frequently cited as a possible analogy to justify the legality of the Albania model. However, as repeatedly emphasized by all parties, the two schemes are fundamentally different from a legal perspective. The Commission’s proposed reform of the notion of a ‘return decision,’ which provides the legal basis for return hubs, presupposes a formal removal decision to a country other than the state of origin with which an agreement exists. In that context, the transfer constitutes an actual return under EU law (discussed, in this blog, here). By contrast, the Albania model does not involve a removal but a temporary relocation of the detention stage of the return procedure under the full jurisdiction of the Member State, prior to any formal removal to the individual’s country of origin or habitual residence. Consequently, it cannot be assimilated to the return hub framework.

Similarly, the Albania model cannot be framed within the reformed notion of a safe third country without a connection requirement (discussed, in this blog, here), as no inadmissibility decision is taken in Italy based on Albania’s acceptance of the asylum application. Instead, the procedure merely relocates the processing of the asylum claim to Albania while keeping the substantive application under Italian jurisdiction.

This distinction highlights that equating the Albania model with return hubs or safe third country schemes overlook its unique legal and procedural features. It underscores the complexity of assessing its compliance with EU law and why the pending preliminary rulings are essential to clarify the boundaries and practical limits of this peculiar migration management scheme.

Ana Carolina Ribeiro Alves (master’s student in Administrative Law at the School of Law of the University of Minho)

From 2024 to today: amendment of Regulation (EU) 2021/1119 (European Climate Law)

On 5 March 2026, the Council formally adopted the amended European Climate Law, introducing a binding intermediate climate target for 2040 of 90% of reduction in net greenhouse gas emissions compared to 1990 levels. It entered into force 20 days after its publication in the Official Journal of the European Union, applying directly in all EU countries. This milestone strengthens the EU’s trajectory towards climate neutrality. The present article provides a timeline of key political and legislative events that led to the adoption of this amendment.

Regulation (EU) 2021/1119[1] established the EU’s climate targets for 2030 and 2050, creating the framework known as the European Climate Law. Article 4(3) requires the setting of an intermediate Union-wide climate target for 2040 to provide Member States with predictability and a clear transition pathway.

On 6 February 2024,[2] the Commission published a Communication on the 2040 EU climate target, outlining a path from the agreed 2030 intermediate goal. Informed by the scientific advice of the European Scientific Advisory Board on Climate Change and through a detailed impact assessment, it recommended a 90% net reduction in greenhouse gas emissions by 2040 relative to 1990.

On 2 July 2025,[3] a proposal for the regulation of the European Parliament and of the Council amending the Regulation (EU) 2021/1119 establishing the framework for achieving climate neutrality was published, to review relevant Union legislation to enable the achievement of the 2040 target and the climate neutrality objective. In this context,[4]  the main focus of the review was to ensure that, from 2036 onwards, up to 3% of the Union’s net emissions in 1990 would be covered by high-quality international credits under Article 6 of the Paris Agreement, as official units, each representing one tonne of CO₂- equivalent, consisting of real, additional and verifiable emission reductions or removals of greenhouse gases generated through international cooperation between countries. These units can be transferred from one country to another, allowing the buyer country to count them toward its climate targets, without double-counting and maintaining high environmental integrity.

Moreover, particular attention is paid to the role of domestic permanent carbon removals (processes that involve capturing carbon dioxide from the atmosphere and storing it durably) under the Union Emissions Trading System (EU ETS) in offsetting residual emissions from hard to abate sectors. In addition, the Commission will ensure greater flexibility between sectors and instruments, to support the achievement of targets in an economically advantageous and socially equitable way.

Implementation policies for this target will be guided by solidarity and equity to ensure a just transition for all Member States and citizens. Coordinated Union action enables consideration of varying capacities across Member States and regions, leveraging the single market as a driver of cost-efficient transformation. Such coordination also strengthens the EU’s international climate action, particularly taking into account that the EU and all its Member States have rectified the Paris Agreement. The initiative’s objective was to establish an intermediate Union-level target for 2040 toward 2050 neutrality, as mandated by Article 4(3) of the European Climate Law. A regulation was deemed the most appropriate instrument to preserve consistency with the existing legislative framework.

Essential elements to keep the EU on track for 2050 neutrality include transparent and regular reporting by Member States, robust Commission assessments, and mechanisms to evaluate progress. Before proposing the 2040 target, the Commission considered multiple factors, with equity and solidarity among and within Member States being the main focus. Given the transboundary nature of climate change, a 2040 Union-wide target cannot be sufficiently achieved by Member States alone but can be better realised at Union level due to its scale and effects. The Union may therefore act in accordance with the subsidiarity principle, as laid down in Article 5 TEU.

On 13 November 2025,[5] the Parliament reached a decision and adopted its position stating that it supported the proposal to amend the European Climate Law, in which a binding EU climate target for 2040 of reducing net greenhouse gas emissions by 90% compared to 1990 levels was set. It stressed that looking ahead to the post-2030 period, the Commission would review relevant EU legislation to enable the achievement of both the 2040 target and climate neutrality. As part of this review, the Commission reiterated that, as of July 2025, high-quality international credits under Article 6 of the Paris Agreement may account for up to 5%, rather than 3%, of the Union’s net emissions in 1990. Furthermore, the matter was referred back to the Committee responsible for interinstitutional negotiations.

On 10 February 2026,[6] the European Parliament adopted a legislative resolution on the proposal for a regulation of the European Parliament and of the Council amending Regulation (EU) 2021/1119 establishing the framework for achieving climate neutrality. As part of this review, the Commission will need to ensure, from 2036 onwards, an adequate contribution towards the 2040 climate target of high-quality international credits of up to 5% of 1990 Union net emissions, corresponding to a domestic reduction of net greenhouse gas emissions by 85% compared to 1990 levels, meaning that at least 85% of emissions reductions must be achieved within the European Union.

On 5 March 2026,[7] the Act was formally adopted by the Council, introducing a binding intermediate climate target for 2040 of a 90% reduction in net greenhouse gas emissions. This amended climate law also changes the date for the EU emissions trading system for road transport, building and other sectors (ETS2) to become fully operational in 2028. This marks the final step in the legislative process, establishing directly applicable rules to all EU Member states. Apart from the previous additions, this amended regulation also provides for a review of the European Climate Law every two years.

The overall framework thus comprises a 55% reduction by 2030, 90% by 2040 (relative to 1990), and climate neutrality by 2050.

The background: European Green Deal and European Climate Law

In 2019, the European Commission unveiled the European Green Deal strategy. As the Commission explains,^[8] it responds to an urgent call from citizens, especially young people, for decisive climate action. The European Green Deal comprises a set of strategic initiatives that steer the European Union toward a genuine ecological transition, outlining a comprehensive plan to make the economy, energy, transport, and industry sustainable.

Its overarching goal is to reduce greenhouse gas emissions by at least 55% by 2030 and 90% by 2040, achieving climate neutrality by 2050. The Green Deal^[9] stresses that every policy area must contribute to combating climate change, supporting targeted measures across all sectors. A central objective is to position the EU as the world’s first climate-neutral continent by 2050, while mitigating pollution and restoring healthy natural and ecosystem balances.

Achieving this requires simultaneous drastic emissions cuts, adoption of a circular economy model (reusing, repairing, and recycling products to minimise waste and preserve resources), promotion of cleaner, more sustainable, and energy-efficient industries, nature restoration plans aiming for zero pollution to secure a healthy environment for future generations, more ecological farming practices that protect the environment while delivering healthier and more affordable food, and a just and inclusive transition plan.

Despite being a soft-law instrument, the European Green Deal gave the political setting so a constitutional grounding could be presented and adopted in the European Union, acting as a complex and ambitious strategy based on a wide range of initiatives, namely (but not limited to) those that demand great investment on climate and environmental action. Insofar, the European Climate Law (Regulation (EU) 2021/1119 of 30 June 2021) codifies the Deal’s objectives, targeting a sustainable Europe that attains climate neutrality by 2050. Its primary focus remains the reduction of greenhouse gas emissions by at least 55% by 2030 relative to 1990. According to the Commission,^[10] its core aims are to set a long-term strategic direction toward 2050 neutrality, establish an ambitious 2030 EU target, creating a robust progress-monitoring system, providing predictability for investors and economic actors, and ensuring the transition to neutrality is irreversible.

It includes in its key elements a legally binding net-zero greenhouse gas emissions target by 2050. EU institutions and Member States are required to adopt necessary measures at both Union and national levels, prioritising equity and solidarity among Member States. At the national level, this is implemented by administrative bodies functioning as a “functionally European public administration,” guided by core EU principles such as sincere cooperation.

In this regard, and bearing in mind the new version of the European Climate Law and its potential to harmonise the applicable legal framework, this can be seen both as a consequence of the principle of sincere cooperation – which has provided the appropriate setting for the European Union to regulate this environmental dimension – and as a further instrument for actively reaffirming the need for further integration. The principle of sincere cooperation – laid down in Article 4(3) TEU – entails interinstitutional coordination, while imposing significant obligations on Member States: they must adopt all appropriate general and specific measures to ensure compliance with EU primary and secondary law obligations, thereby giving effect to the European Climate Law, while securing an European role in achieving this normative act’s goals.

Conclusion: diluting ambition

While the formal adoption of a 90% net greenhouse gas emissions reduction target for 2040 appears to strengthen the EU’s climate framework, the simultaneous extension of the timeline and the introduction of international carbon credits under Article 6 of the Paris Agreement reveals a concerning dilution of ambition. In practice, this mechanism lowers the required domestic reductions allowing Member States to compensate for residual emissions through international credits rather than accelerating the necessary deep decarbonisation in sectors that are most challenging. Such flexibility raises serious doubts about whether the EU is truly on track to meet its 2030 targets or if it is merely postponing difficult and political and economic choices through extended deadlines and creative accounting. This ultimately relies on the principle of sincere cooperation (Article 4(3) TEU), which demands loyal and effective action from both Member States and EU institutions. However, if this principle is used as an excuse to accommodate delays and compensate for deficiencies through international credits rather instead of enforcing rigorous domestic efforts, it may undermine the credibility and overall effectiveness of the European Climate Law, casting doubt on the EU’s leadership in global climate action.

[1] Regulation (EU) 2021/1119 of the European Parliament and of the Council of 30 June 2021 establishing the framework for achieving climate neutrality and amending Regulations (EC) No 401/2009 and (EU) 2018/1999 (‘European Climate Law’), available at https://eur-lex.europa.eu/eli/reg/2021/1119/oj/eng, accessed on 11.03.2026.

[2] Communication from the Commission to the European parliament, theCouncil, the European Economic and Social Committee and the Committee of the Regions, Securing our future Europe’s 2040 climate target and path to climate neutrality by 2050 building a sustainable, just and prosperous society, available at https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:52024DC0063, accessed on 11.03.2026.

[3] Proposal for a Regulation of the European Parliament and of the Council amending Regulation (EU) 2021/1119 establishing the framework for achieving climate neutrality, available at https://www.europarl.europa.eu/RegData/docs_autres_institutions/commission_europeenne/com/2025/0524/COM_COM(2025)0524_EN.pdf, accessed on 11.03.2026.

[4] European Parliament, Framework for achieving climate neutrality, 2025/0524(COD), 2 July 2025, available at https://oeil.europarl.europa.eu/oeil/en/document-summary?id=1826240, accessed on 11.03.2026.

[5] European Parliament, Framework for achieving climate neutrality.

[6] European Parliament, Framework for achieving climate neutrality.

[7] Council of the EU, “2040 climate target: Council gives final green light”,Press Release, 5 March 2026, available at https://www.consilium.europa.eu/en/press/press-releases/2026/03/05/2040-climate-target-council-gives-final-green-light/, accessed on 11.03.2026

[8] European Commission, “The European Green Deal. Striving to be the first climate-neutral continent”, available at https://commission.europa.eu/strategy-and-policy/priorities-2019-2024/european-green-deal_en, accessed on 07.01.2026.

[9] Council of the EU, “European Green Deal”, available at: https://www.consilium.europa.eu/en/policies/european-green-deal/, accessed on 07.01.2026.

[10] European Commission, “European Climate Law”, available at https://climate.ec.europa.eu/eu-action/european-climate-law_en, accessed on 07.01.2026.

Picture credit: by Markus Spiske on pexels.com. 

The post How Péter Magyar and Tisza remade the Hungarian party system first appeared on LSE European Politics.

The post How Hungary’s Tisza Party won everything, everywhere, all at once first appeared on LSE European Politics.

The post It’s time for a more pragmatic EU-China trade relationship first appeared on LSE European Politics.

The post What should we expect from the next Hungarian government? first appeared on LSE European Politics.

Ana Filipa Ribeiro (master’s student in European Union Law at the School of Law of University of Minho and ENDE Research Grant Holder – UMINHO/BIM/2026/33)

Regulation (EU) 2023/1114 on Markets in Crypto-assets (MiCAR)[1] is the European Union’s first comprehensive framework for crypto-assets that fall outside existing financial-services legislation,[2] designed to harmonise rules across the internal market, while pursuing objectives traditionally associated with public regulation, including investor and consumer protection, market integrity and financial stability.[3] MiCAR does so in large part by placing crypto-asset service providers (CASPs) at the centre of its governance architecture. A CASP[4] is a legal person or other undertaking that is authorised and supervised as an intermediary that provides one or more crypto-asset services to clients on a professional basis, including custody and administration of crypto-assets, operation of trading platforms, exchange, execution of orders and transfers on behalf of clients.[5] In practice, CASPs set and enforce platform rules, monitor activity, restrict access, freeze, or limit the movement of assets, suspend trading and delist tokens, often in real time and on the basis of risk assessments that combine regulatory obligations and internal policies.[6] It is therefore plausible to say that these entities go beyond acting as technical conduits and participate in a form of private ordering with structurally quasi-public effects, a claim developed below.

Because CASPs control the infrastructure through which most users access crypto markets, their decisions can function as immediate constraints on market access and on effective enjoyment of asset-related interests. Crucially, these decisions often give rise to dispute’s origin, by unilaterally altering a client’s position through measures such as freezing assets, restricting transfers, suspending trading, or delisting a token.[7] The resulting conflict is then typically channelled into an internal process designed and administered by the CASP itself, with the provider acting simultaneously as rule-maker, investigator, decision-maker, and initial reviewer.[8] This concentration of functions is a familiar source of legitimacy concerns in public governance, yet it has received limited conceptual attention in the MiCAR context.

A useful way to sharpen these legitimacy concerns is to rely on the functional notion of quasi-public powers developed in the academic literature.[9] This is not to say that private actors become public authorities, but that, where public entities have a limited ability to interfere effectively in digital environments, certain private operators end up exercising powers that can be enforced autonomously, sometimes automatically, through the infrastructure they control and for which users often find few meaningful means of recourse, with redress mechanisms largely shaped by the provider itself. The idea of quasi-public power by digital platforms or services can be structure through a threefold that mirrors classic public functions:[10] private operators can exercise (1) a quasi-legislative function by unilaterally setting the rules of the environment; (2) a quasi-executive function by enforcing those rules through restrictions; and (3) a quasi-judicial function by resolving conflicts on a case-by-case basis, including through balancing competing interests. Once rulemaking, enforcement and dispute resolution are concentrated in a single actor, legitimacy questions usually associated with public governance become structurally relevant, even when the legal form remains private. 

Read against MiCAR’s architecture, the parallel becomes difficult to ignore. CASPs effectively perform a quasi-legislative role when they define the operational rules and participation conditions that structure access to their services and to the liquidity they intermediate. They perform a quasi-executive role when those rules are enforced through self-executing measures inside the controlled infrastructure, such as freezes, transfer restrictions, suspensions and delistings. And they perform a quasi-judicial role to the extent that adverse decisions are contested through internal channels that the CASP administers, a design MiCAR expressly assumes by requiring CASPs to maintain complaints-handling procedures,[11] which are then further specified in the delegated technical standards on complaints handling.[12] In this framing, the quasi-public exercise of powers by CASPs does not describe a moral judgment about CASPs, but instead a governance structure. Once MiCAR places these intermediaries at the centre of market access and user protection, the central question is no longer whether such powers exist in practice, they clearly do. The question is what renders the exercise of those powers legitimate within an EU regulatory programme that pursued public objectives, while operating through private infrastructures.

By design, MiCAR ties the exercise of CASP power to a set of public-regulatory aims[13] and to specific organisational and conduct constraints.[14] This suggests that CASPs are not left to operate purely as private contractors, as they are authorised and supervised market actors, subject to duties that reach beyond pure disclosure and into how they must behave when dealing with clients. In particular, MiCAR requires CASPs to act honestly, fairly, and professionally in the best interests of clients and it imposes governance and conflicts-of-interests requirements that implicitly recognise the risk of self-serving decision-making in a setting where the provider both runs the infrastructure and benefits from it.[15]

Yet, these conduct and governance duties only perform a genuine legitimacy function if users have practical ways to react when a CASP takes an adverse decision. In practice, this may give rise to concerns regarding contestability and exit. Contestability, in this regard, covers whether a user can meaningfully contest a restrictive measure, obtain reasons, and secure a review capable of changing the outcome, while exit concerns whether the user can realistically move to another provider or whether infrastructure control and ecosystem dynamics make departure illusory.

Contestability under MiCAR starts inside the CASP, as the Regulation requires CASPs to maintain effective and transparent complaints-handling procedures and to investigate complaints in a timely and fair manner, while the delegated technical standards specify how this channel must operate in practice, including communication of outcomes and procedural discipline.[16] MiCAR also provides for complaint channels at the level of national competent authorities, with ESMA publishing links to those procedures.[17] Even so, the first response to a freeze, restriction, suspension, or delisting will typically be processed within the provider that imposed it, which keeps informational asymmetry and institutional self-review at the centre of the dispute. This is precisely where the EU principle and fundamental right to effective judicial protection becomes relevant. As a general principle of EU law anchored in Article 19(1) of the Treaty of European Union and developed by the Court of Justice, it presupposes that individuals have access to remedies and procedural guarantees that secure, at minimum: (1) the right of action, (2) the right to an independent and impartial tribunal, (3) the rights of defence, and (4) access to justice for those who lack sufficient resources.[18] In parallel, Article 47 of the Charter of Fundamental Rights of the European Union guarantees an effective remedy and a fair trial when EU-law rights and freedoms are at stake. In this perspective, procedural regularity cannot be reduced to the formal existence of a route to court, as it requires that the procedural set-up is materially capable of enabling the person affected to defend their rights effectively, in line with the Rule of Law commitments of Article 2 TEU. Read against this light, contestability under MiCAR cannot be viewed merely as a consumer service expectation. As a matter of fact, contestability becomes one of the practical conditions through which effective judicial protection is realised. If the decisive stage of review takes place internally, a user’s ability to obtain a meaningful remedy later depends on whether they can understand the measure they are facing (what was decided, on what basis, with what effects and for how long). Without intelligible reasons and clear terms, the right to be heard and the right to defence risk becoming hollow, because the affected person cannot frame a challenge, gather relevant evidence, or assess whether escalation to a competent authority or a court is warranted.

Exit is the second constraint. MiCAR does not create an EU-wide “ban” that automatically follows a customer from one CASP to another, nor does a delisting on one platform automatically remove an asset from the market as such. In practice, however, exit can be constrained when the initial CASP controls custody and transfers or when multiple platforms converge on similar risk assessments and replicate restrictions across the ecosystem. Therefore, exit constraints could also have a competition dimension. When liquidity and user access are concentrated in a small number of CASPs,[19] restrictive decisions may raise switching costs and, in cases of market power or coordinated behaviour, could engage EU competition-law concerns.

Once these two correctives are placed next to MiCAR’s reliance on intermediaries, the legitimacy question becomes sharper. If contestability is mostly internal and weak and if exit is practically constrained, then CASPs’ capacity to set rules, enforce them through infrastructure and review disputes within their own procedures begins to resemble a closed governance loop.

Nevertheless, this architecture is understandable from a regulatory standpoint. MiCAR is meant to create a harmonised EU framework that supports innovation and fair competition while pursuing objectives such as a high level of protection for retail holders and the integrity of crypto-asset markets.[20] Rather than attempting to regulate protocols as such (something that is hardly possible given the decentralised nature of the technology), the EU governs the market where accountability, supervision and enforcement are realistically available.

On the other hand, it is reasonable to recognise that now decentralisation often describes the technology itself more than the dominant patterns of use. The original Bitcoin white paper framed the core promise as enabling payments “directly from one party to another without going through a financial institution”.[21] Contemporary market practice frequently departs from that ideal, as many of the largest and most popular exchanges function as centralised intermediaries and, in practice, can negate the aim of decentralisation.[22] In that environment, relying on CASPs is less a rejection of decentralised technology than a pragmatic response to the fact that, for most users, access to crypto markets is mediated, concentrated and operationally governed through private infrastructures.

That choice of architecture, even if understandable, does not make the legitimacy question disappear. Once MiCAR relies on CASPs as the main points of access to liquidity, custody, execution and market participation, the moment when a CASP freezes assets, restricts transfers, suspends trading, or delists a token cannot be seen as a routine contractual event. Rather, it is experienced as a market-shaping intervention that is immediately effective because it is executed through infrastructure control and that often becomes the very source of the dispute the user is then forced to navigate.

If quasi-public powers are the possible solution, the corresponding responsibility cannot be ignored. MiCAR already sketches the legitimacy constraints that should travel with that power, namely conduct duties that exclude arbitrary treatment, organisational and conflicts-of-interest requirements that address self-serving incentives, operating rules that discipline platform intervention and mandatory complaints-handling that structures the first line of contestation. These are the conditions that keep the model from collapsing into a closed circuit where the same entity sets the rule, enforces it, and validates its own enforcement. In practice, this means that the credibility of MiCAR’s regulatory settlement depends on whether high-impact measures are intelligible and contestable. As the Rule of Law imposes, users must be able to understand what happened and why, restrictive measures should be time-bounded rather than indefinite by inertia, internal review must be capable of changing outcomes rather than merely confirming them and decisions should remain traceable for supervisory scrutiny and, where necessary, judicial review.

[1] For a complementary analysis of the MiCA Regulation, developed from a different perspective, see our post: Ana Filipa Ribeiro, “New digital manifestations of financial services and european integration: what benefits for the european citizen,” Official Blog of UNIO – Thinking and Debating Europe, April 25, 2024, accessed March 5, 2026, https://officialblogofunio.com/2024/04/25/new-digital-manifestations-of-financial-services-and-european-integration-what-benefits-for-the-european-citizen/.

[2] Regulation (EU) 2023/1114 of 31 May 2023 on markets in crypto-assets and amending Regulations (EU) No 1093/2010 and (EU) No 1095/2010 and Directives 2013/36/EU and (EU) 2019/1937, CELEX:32023R1114, recital 6.

[3] European Securities and Markets Authority (ESMA), “Markets in Crypto-Assets Regulation (MiCA),” ESMA, last update February 23, 2026, accessed March 5, 2026, https://www.esma.europa.eu/esmas-activities/digital-finance-and-innovation/markets-crypto-assets-regulation-mica.

[4] The definition can be found in Article 3(1)(15) of MiCAR.

[5] MiCAR clarifies that it applies to crypto-asset services and activities performed, provided, or controlled by natural or legal persons and certain other undertakings including when part of such activities or services is performed in a decentralised manner. Where crypto-asset services are provided in a fully decentralised manner without any intermediary, they should not fall within its scope. See Regulation (EU) 2023/1114, recital 22. Accordingly, purely self-custodial tools that do not safeguard or control clients’ assets and/or keys on their behalf and do not otherwise provide a listed crypto-asset service (as hardware wallets or non-custodial software wallets) will generally fall outside the CASP category, although the classification depends on the concrete functionality and business model.

[6] Regulation (EU) 2023/1114, article 76.

[7] Regulation (EU) 2023/1114, article 76(1)(b), (f), (g).

[8] Regulation (EU) 2023/1114, article 71.

[9] Giovanni De Gregorio, “From constitutional freedoms to the power of the platforms: protecting fundamental rights online in the algorithmic society,” European Journal of Legal Studies 11(2), (Spring 2019): 65-103, accessed March 5, 2026, https://cadmus.eui.eu/server/api/core/bitstreams/17922b19-9645-5935-901a-6b4b942e2682/content. See also Miguel Pereira, “Mapping the values of digital constitutionalism: guiding posts for digital Europe?”, UNIO – EU Law Journal 10 (2), (December 2024): 70-90, accessed March 2, 2026, https://revistas.uminho.pt/index.php/unio/article/view/6045/6907.

[10] De Gregorio, “From constitutional freedoms to the power of the platforms”, 85-89. Pereira, “Mapping the values of digital constitutionalism”, 75.

[11] Regulation (EU) 2023/1114, article 71.

[12] Commission Delegated Regulation (EU) 2025/294 of 1 October 2024 supplementing Regulation (EU) 2023/1114 of the European Parliament and of the Council with regard to regulatory technical standards specifying the requirements, templates and procedures for the handling of complaints by the crypto-asset service providers, Official Journal of the European Union L 2025/294 (February 13, 2025), accessed March 5, 2026

[13] An EU framework is needed “to foster innovation and fair competition” while ensuring “a high level of protection of retail holders” and “the integrity of crypto-asset markets”, as expressed in Regulation (EU) 2023/1114, recital 6.

[14] Regulation (EU) 2023/1114, articles 66, 67, 68, 70, 71 and 72.

[15] Regulation (EU) 2023/1114, article 66.

[16] Commission Delegated Regulation (EU) 2025/294, article 1(2). 

[17] Regulation (EU) 2023/1114, article 108.

[18] Joana Covelo de Abreu, “Princípio da tutela jurisdicional efetiva”, in Enciclopédia da União Europeia, ed. Ana Paula Brandão et al. (Lisbon: Petrony, 2017), 329-32.

[19] On market concentration, ESMA reports that despite the large number of crypto-assets, both market capitalisation and trading activity remain significantly concentrated. See Malena Calissano, Filippo Giuglini, and Paul Reiche, “Crypto assets: market structures and EU relevance”, European Securities and Markets Authority TRV Risk Analysis, ESMA50-524821-3153, April 10, 2024, 9-10, accessed March 10, 2026, https://www.esma.europa.eu/sites/default/files/2024-04/ESMA50-524821-3153_risk_article_crypto_assets_market_structures_and_eu_relevance.pdf.

[20] It will not be discussed whether one agrees or disagrees with this approach, as that is not the scope of this article.

[21] Satoshi Nakamoto, “Bitcoin: a peer-to-peer electronic cash system”, October 31, 2008, 1, accessed March 9, 2026, https://bitcoin.org/bitcoin.pdf.

[22] Parma Bains, Arif Ismail, Fabiana Melo, and Nobuyasu Sugimoto, “Regulating the crypto ecosystem: the case of unbacked crypto assets,” FinTech Notes no. 2022/007 (Washington, DC: International Monetary Fund, September 2022), 18, accessed March 11, 2026, https://doi.org/10.5089/9798400221361.063.

Picture credit: by Alesia Kozik on pexels.com.

The post Why don’t all immigrants support progressive parties? first appeared on LSE European Politics.

Gonçalo Martins de Matos (PhD candidate in Public Legal Sciences at the School of Law of the University of Minho | Junior Researcher at JusGov | Member of the Editorial Support of this blog)

As the Hungarian legislative elections approach, we are reminded of what is at stake for the whole of the European Union (EU). Since 2010, the Hungarian State’s democratic and Rule of Law standards have backslid, turning this Eastern European State in a de facto illiberal democracy,[1] dominated by Viktor Órban and his Fidesz party. Órban’s rule has been generally uncontested, even with European institutions increasingly drawing attention to Hungary’s severe democratic decline.[2] However, his grip on the Hungarian State has been facing rising challenges, such as several infringement procedures[3] and the activation of the Rule of Law conditionality mechanism.[4] The 2024 European Parliament elections gave challenger Péter Magyar the upward momentum to hinder Órban’s illiberal agenda and shake the foundations of his firm grasp on Hungary’s legal political system.[5]

The possibility of halting or reversing Hungary’s democratic decline is exponentiated by an infringement procedure pending before the Court of Justice of the European Union (CJEU). Case C- 769/22 Commission v. Hungary[6] is currently awaiting the CJEU’s ruling. However, when the Advocate General’s Opinion was published in June 2025,[7] we soon realised that the implications may go beyond the Hungarian case and help found what has been referred to as the justiciability of Article 2 of the Treaty on European Union (TEU). In summary, the legal argument is that the values enshrined in Article 2 TEU create concrete legal obligations for the Member States. If Member States fail to fulfil those obligations, their non-compliance can be used as grounds for launching infringement procedures, well within the ordinary competences of the CJEU. If the upcoming judgment eventually adopts the justiciability argument, we may be looking at a whole new phase of the constitutionality control carried out by the CJEU, keeping alive the aspiration of a society in which pluralism, non-discrimination, tolerance, justice, solidarity and equality between women and men prevail.

However, given that the CJEU’s contributions to EU law are built from each individual case to by its case-law, the basis of the qualitative leap to the justiciability of Article 2 TEU can be detected in a CJEU judgment from 18 December 2025. Case C-448/23 Commission v. Poland[8] concerned two 2021 rulings by the Polish Constitutional Court (PCC) that essentially intended to challenge the primacy of EU law. In the 14 July ruling, the PCC held that the CJEU interim measure of suspending the Disciplinary Chamber of the Polish Supreme Court imposed ultra vires obligations on the Polish Republic, thus breaching the principle of conferral. In the 7 October ruling, the PCC claimed that the interpretation of Articles 2 and 19 TEU made by the CJEU in another infringement ruling[9] was incompatible with the Polish Constitution, thus undermining Poland’s national constitutional identity.

Before we proceed to the findings of the CJEU’s judgment, we must contextualise these PCC rulings in the broader Rule of Law backsliding that has plagued some EU Member States since 2014. Rule of Law backsliding is defined by a gradual capture of a State’s legal and political systems by a party in power, provoking severe setbacks in areas that previously seemed protected, such as fundamental rights, legality or transparency, to name some. Since the rise to power of the illiberal party PiS in 2015, Poland has been immersed in a constitutional crisis, having merely slowed down in 2023, with the parliamentary defeat of the illiberal government. Through partisan appointments of judges to the Supreme and Constitutional Courts, the PiS-led government effectively weaponised judicial bodies to forward their illiberal agenda, with lasting impact to this day.

Against this backdrop, EU institutions have been tackling Rule of Law backsliding, and, despite an initially shaky and timid reaction, there have been significant advancements in enforcing the Rule of Law, especially through the reconfiguration of the infringement procedure to tackle systemic breaches of the fundamental values of the EU.[10] Especially since the 2019 Commission v. Poland judgment,[11] in which the CJEU “tackled the issue of principle at the heart of the matter: adherence to the Rule of Law via honouring judicial independence and irremovability”.[12] The CJEU judgment we are analysing is the latest in the sequence initiated with this 2019 landmark ruling.

With all of this context, and taking into account the findings of the PCC rulings, the CJEU rose to the task and reinforced basic principles of EU law, while simultaneously furthering the innovative aspects inaugurated with the 2019 Commission v. Poland judgment. In this 2025 judgment, the CJEU has ruled that the binding force of EU law or its own judgments cannot be undermined by either the principle of conferral or national identity, expressly rejecting national constitutional identity as a justification for disapplying CJEU case law. The CJEU argued that, since the values enshrined in Article 2 TEU form part of the constitutional identity of the Union as a common legal order, to which the Member States have adhered, their legal concretisation cannot be regarded as an intrusion upon national identity.[13] Furthermore, the CJEU also reinforces the principle of non-regression, reminding that a Member State “cannot, therefore, amend its legislation, or indeed its case-law, in such a way as to bring about a reduction in the protection”[14] of fundamental values, with particular emphasis on the Rule of Law.

Recalling that the principle of conferral demands that each EU institution act only within the limits of the competences conferred upon it by the Member States, the CJEU emphasised that the authority to determine whether those limits have been exceeded or not belongs only to the CJEU itself.[15] If a national court considers that a provision of EU law is invalid on the grounds that it exceeds the Union’s competences, or that it encroaches national identity, it must refer a question to the CJEU for a preliminary ruling, according to Article 267 of the Treaty on the Functioning of the European Union (TFEU), “the [CJEU] alone having jurisdiction to declare an EU act invalid”.[16]

With the judgment sub judice, the CJEU, by reinforcing the supremacy of its authority on the interpretation of EU law, consequently reaffirms the principle of the primacy of EU law. However, it goes further than that. By asserting that the values enshrined in Article 2 TEU are not merely a statement of policy guidelines or intentions but “are given concrete expression in principles containing legally binding obligations for the Member States”,[17] the CJEU is paving the way to enable the justiciability of the values contained in Article 2 TEU.[18] In addition, by asseverating that the only authority with the competence of knowing of the interpretation or validity of EU law is the Court itself, through the reference for a preliminary ruling procedure, the CJEU once again strengthens the principle of effective judicial protection.

From what we have just analysed, the CJEU, by reinvigorating fundamental aspects of legality within the EU, is laying the groundwork for the effective judicial protection of the very values that define the EU as a Union of Law. In addition to this, if the CJEU adopts, in the impending ruling on Case C-769/22 Commission v. Hungary, some of the basic arguments provided by the above-mentioned Advocate General’s Opinion, in theory, the CJEU will have the ordinary competence of declaring the infringement of the fundamental values of the EU. The legal ramifications of both developments in the infringement procedure are examined in greater detail in an article written by us to be published in the next issue of UNIO – EU Law Journal.

[1] On rule of law backsliding, see, among others, Carlos Closa and Dimitry Kochenov, “Part I. The case for EU reinforced oversight in four questions” in Carlos Closa, Dimitry Kochenov and JHH Weiler, EUI Working Paper RSCAS 2014/25: Reinforcing Rule of Law Oversight in the European Union, no. 3 (2014): 7, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2404260, Laurent Pech and Kim Lane Scheppele, “Illiberalism within: rule of law backsliding in the EU”, Cambridge Yearbook of European Legal Studies, vol. 19 (2017): 7, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3009280, and Gonçalo Martins de Matos, “Defending the Rule of Law in the European Union: legal and political approaches”, UNIO – EU Law Journal, vol. 11, no. 2 (2025): 102-104, https://doi.org/10.21814/unio.11.2.7018.

[2] From the well-known “Tavares” Report (European Parliament, Report A7-0229/2013, https://www.europarl.europa.eu/doceo/document/A-7-2013-0229_EN.html) to the successive Rule of Law Reports (European Commission Reports on the situation of the Rule of Law in the EU, available at https://commission.europa.eu/strategy-and-policy/policies/justice-and-fundamental-rights/upholding-rule-law/rule-law/annual-rule-law-cycle_en#:~:text=The%20European%20Rule%20of%20Law%20Mechanism%20provides%20a,institutions%20to%20contribute%20in%20accordance%20with%20their%20).

[3] See, among others, Kim Lane Scheppele, Dimitry Kochenov and Barbara Grabowska-Moroz, “EU values are law, after all: enforcing EU values through systemic infringement actions by the European Commission and the Member States of the European Union”, Yearbook of European Law, vol. 39, no. 3 (2020): 42-44, https://academic.oup.com/yel/article/doi/10.1093/yel/yeaa012/6064852?login=true.

[4] We have written about this subject in Gonçalo Martins de Matos, “On the triggering of the EU’s conditionality mechanism: what has been done and what could follow”, The Official Blog of UNIO – Thinking and Debating Europe, 3 February 2023, https://officialblogofunio.com/2023/02/03/on-the-triggering-of-the-eus-conditionality-mechanism-what-has-been-done-and-what-could-follow/.

[5] Vincenzo Genovese and Sandor Zsiros, “MEPs rally behind Magyar in Hungarian elections despite ideological divides”, Euronews, 30 March 2026, https://www.euronews.com/my-europe/2026/03/30/meps-rally-behind-magyar-in-hungarian-elections-despite-ideological-divides.

[6] Court of Justice of the European Union, Case C-769/22, European Commission v. Hungary, available at https://curia.europa.eu/juris/liste.jsf?lgrec=fr&td=;ALL&language=en&num=C-769/22&jur=C.

[7] We have written about the Advocate General’s Opinion on this case, analysing its main arguments, in Gonçalo Martins de Matos, “Protecting fundamental rights through the infringement procedure: Advocate General’s Opinion in CJEU Case C-769/22 ‘Commission v. Hungary’”, The Official Blog of UNIO – Thinking and Debating Europe, 16 June 2025, https://officialblogofunio.com/2025/06/16/protecting-fundamental-rights-through-the-infringement-procedure-advocate-generals-opinion-in-cjeu-case-c-769-22-commission-v-hungary/.

[8] Judgment CJEU European Commission v. Republic of Poland, 18 December 2025, Case C‑448/23, ECLI:EU:C:2025:975.

[9] Judgment CJEU European Commission v. Republic of Poland, 15 July 2021, Case C‑791/19, ECLI:EU:C:2021:596.

[10] For a comprehensive explanation of the concept of systemic infringement procedure, see Kim Lane Scheppele, “What can the European Commission do when Member States violate basic principles of the European Union? The case for systemic infringement actions”, Verfassungsblog, November 2013, 1, https://verfassungsblog.de/wp-content/uploads/2013/11/scheppele-systemic-infringement-action-brussels-version.pdf.

[11] Judgment CJEU European Commission v. Republic of Poland, 24 June 2019, Case C-619/18, ECLI:EU:C:2019:531.

[12] Scheppele, Kochenov and Grabowska-Moroz, “EU values are law, after all”, 45.

[13] Emmanuel Renoud Delarque, “Ending constitutional resistance? The CJEU’s rejection of ultra vires review and national identity doctrines”, European Law Blog, 11 February 2026, https://www.europeanlawblog.eu/pub/kdmsojzk/release/1. See also recitals 177 to 195 of Judgment Commission v. Poland (C-448/23).

[14] Judgment Commission v. Poland (C-448/23), recital 179.

[15] Judgment Commission v. Poland (C-448/23), recital 213.

[16] Judgment Commission v. Poland (C-448/23), recital 230.

[17] Judgment Commission v. Poland (C-448/23), recital 177.

[18] For more on Judgment Commission v. Poland (C-448/23), see Mariusz Muszyński, “An error or ultra vires action. The CJEU’s judgement C-448/23 against Poland”, Central European Association for Comparative Law, 27 January 2026, https://www.ceaclaw.org/post/an-error-or-ultra-vires-action-the-cjeu-s-judgement-c-448-23-against-poland, Jakub Jaraczewski and Laurent Pech, “Kangaroo courts and EU law: on the Court of Justice’s judgment in Commission v. Poland”, Verfassungsblog, 7 January 2026, https://verfassungsblog.de/commission-v-poland/ (with some criticism) and Wojciech Sadurski, “The CJEU versus the constitutional tribunal in Poland: On the CJEU’s judgment in Case C-448/23 (European Commission v. Republic of Poland)”, Verfassungsblog, 20 December 2025, https://verfassungsblog.de/the-cjeu-versus-the-constitutional-tribunal-in-poland/ (with some criticism).

Picture credit: by Son Tung Tran on pexels.com.

The post Finland’s “Orpo reforms” have failed to deliver first appeared on LSE European Politics.

Dr. Izabella Majcher, Independent Consultant

Photo credit: Alamy stock photo

The current revision of the EU rules on return has progressed rapidly since the European Commission published its proposal for a Return Regulation on 11 March 2025 (discussed here and here). The Council of the European Union adopted its “general approach” to the proposal on 8 December 2025 (discussed here). In addition to numerous critical opinions from civil society organisations (see, for instance, here, here and here), on 26 January 2026, in an unprecedented step, 16 UN experts addressed a joint letter to both the Commission and the Council expressing serious concerns about the proposal. The measures attracting particular criticism included weakened procedural safeguards; expanded immigration detention; the de-prioritisation of “voluntary” departure; increased cooperation obligations, penalties and restrictions; risks of racial and religious profiling; limitations on access to socio-economic rights; insufficient protection of persons in vulnerable situations; and the introduction of “return hubs”. To date, neither institution has responded to the UN experts.

On the side of the European Parliament, within the Committee on Civil Liberties, Justice and Home Affairs (LIBE), the rapporteurship for this file was assigned to the Renew political group. The Rapporteur presented his draft report on 30 October 2025, which attracted over 2,200 amendments from political groups. As intra-LIBE negotiations on these amendments proved challenging, shortly before the planned vote, the largest political group, the European People’s Party (EPP), tabled alternative compromise amendments with the support of right-wing and far-right groups (European Conservatives and Reformists, Patriots for Europe and Europe of Sovereign Nations). Holding a majority within LIBE, these groups secured the report's adoption on 9 March 2026. Despite mobilisation by civil society and observations issued by the Council of Europe Commissioner for Human Rights, the Parliament adopted the report in plenary on 26 March 2026, with 389 votes in favour, 209 against, and 32 abstentions.

This blog post provides a brief discussion of the European Parliament’s position on selected key measures proposed by the Commission and may be read alongside the previous analysis.

Detection measures: As a rare instance in which the Parliament prevents the application of coercive measures, it removed the detection measures proposed by the Commission in Art. 6. These measures, aimed at identifying persons staying in an irregular situation for the purpose of their removal, had been criticised as likely to lead to racial profiling.

Return decision and safeguards (Art. 7, 12 and 26–28): The Parliament expanded the possibility, already envisaged in the Commission’s draft, of not indicating the country of return in the return decision. It provides that, where the country of return cannot be determined at the time of issuing the return decision, the decision may provisionally designate none, one or several countries of return. It further clarifies that the country of return may be determined either in the return decision itself or subsequently, in the removal order. Moreover, where a country of removal has not previously been specified in the return decision, or where it differs from the country designated therein, the Member State is required only to notify the person in accordance with national law. In such cases, the issuance of a new return decision or removal order is not required. This approach may prevent individuals from effectively appealing removal to a specific country, as notifications falling short of a formal decision or act may not be subject to review. It therefore risks undermining the principle of non-refoulement, the right to an effective remedy, and the principle of good administration.

The absence of an indication of the country of return in the return decision is highly problematic from a non-refoulement perspective, as it suggests that an individual assessment of the risk of refoulement may not be carried out at that stage of the procedure. Indeed, the Parliament’s report requires Member States to determine the country of return only “prior to carrying out the removal.” In addition, it weakens the risk assessment envisaged in the Commission’s proposal by transforming what appears to be an obligatory assessment into one that is triggered only when the person raises the risk or when the authorities otherwise become aware of it.

These concerns are compounded by a further weakening of remedies. The Parliament’s report explicitly states that an appeal does not have an automatic suspensive effect, while merely preserving the possibility for national law to provide for such an effect.

Mutual recognition of return decisions (Art. 9): The Parliament largely retains the Commission’s provisions on mutual recognition. It adds, however, that not only return decisions but also removal orders may be enforced by another Member State. This raises practical challenges, as only the main elements of the return decision are to be included in the standardised form (European Return Order) made available through the Schengen Information System. Crucially, given the expanded possibility of not indicating the country of return in the return decision, the enforcing Member State may in practice need to carry out a new procedure to issue a return decision. The Parliament further clarifies that Member States are not required to adopt administrative decisions or acts for the purposes of recognition, which may further limit the individual’s ability to challenge enforcement measures.

Voluntary departure (Art. 13): Compared to the Commission’s proposal, the Parliament further deprioritises “voluntary” departure. While the Commission’s draft implicitly allows for a period of zero days for voluntary departure, the Parliament’s report explicitly provides for the possibility of imposing immediate “voluntary” departure. It also makes the granting of a longer period contingent upon a request by the person concerned and removes the non-exhaustive list of circumstances justifying such an extension, including family links and children’s schooling. The Parliament appears to overlook that voluntary departure is not only more favourable to the individual, and consistent with the principle of proportionality, but also serves the interests of the State.

Safeguards pending removal (Art. 14): The Parliament removes the basic rights to which persons are, in principle, entitled during the period of postponement. This list includes the fulfilment of basic needs, respect for family unity, emergency health care and essential treatment of illness, access to the basic education system, and the protection of the special needs of vulnerable persons. These safeguards are laid down in the current Directive (Art. 14(1)), and the Commission’s proposal slightly expanded this list. They also reflect human rights obligations, including the rights to education, health, and basic subsistence.

Monitoring of removal (Art. 15): The detailed provisions on return monitoring included in the Commission’s proposal have been significantly curtailed. Notably, the Parliament removed key elements relating to the powers and capacities of the independent monitoring mechanism, including the requirement to ensure appropriate means, autonomy in action, and a mandate to communicate substantiated allegations of failures to respect fundamental rights to the competent national authorities.

Detention (Art. 29–35): The Parliament’s report retains the Commission proposal’s deletion of the current requirement that detention may be imposed only where no sufficient but less coercive measures can be applied in a specific case (Art. 15(1)). It also expands the grounds for detention to include non-compliance with cooperation duties. Given this extensive list of grounds, detention risks becoming a regular measure, in violation of the principles of necessity and proportionality inherent in the right to liberty. The maximum detention period of two years (12 months + 12 months) in a given Member State is maintained, representing a significant increase from the current 18-month limit (Art. 15(5)-(6)). In contrast to the Commission’s draft, the review of detention would no longer need to be concluded within 15 days. On a positive note, the Parliament introduced, in the preamble (§32), a provision stating that children, as a rule, should not be detained.

As regards detention conditions, the Parliament retains the limitations proposed by the Commission and, in some respects, further amplifies them. Compared to the current directive (Art. 16), the proposed rules significantly lower standards governing detention conditions and treatment. Detention may take place not only in specialised facilities but also in designated sections of other facilities, and where prisons are used, migrants are to be kept separated from ordinary prisoners only where possible. Temporary measures may be adopted to expand capacity in exceptional return situations. Access to open-air spaces, introduced by the Commission, may be temporarily restricted where necessary and proportionate to ensure the proper functioning of detention facilities. Contact with the outside world will be subject to security and administrative considerations. Families without children will not be entitled to separate accommodation ensuring privacy.

Entry ban (Art. 10–11): The Parliament further broadens the already expanded scope of entry bans proposed by the Commission. It removes the maximum duration of 10 years for entry bans (compared to the current maximum of 5 years in cases not involving security threats under Art. 11(2)) and also eliminates any maximum period for persons considered to pose security risks. In addition, the Parliament renders less explicit the circumstances in which an entry ban may be withdrawn, suspended, or shortened.

It also removes the safeguards proposed by the Commission for entry bans imposed when a person’s irregular stay is detected at exit, including the requirement for justification based on the specific circumstances of the individual case, respect for the principle of proportionality, and the rights of defence. Furthermore, the Parliament’s report introduces additional situations in which entry bans may be imposed in the context of departure, such as where a person leaves the territory without a return decision having been issued, or departs before such a decision is adopted. Entry bans imposed on persons seeking to leave the territory without completing a return procedure are counterproductive, disproportionate, and disregard the requirement of an individual assessment.

Return hubs (Art. 4(3) and 17): The Parliament’s report provides that the agreement or arrangement forming the basis for return hubs may be concluded not only by a Member State but also by the Union. It also weakens notification obligations towards the Commission and removes families with children from the category of persons exempted from removal to return hubs, leaving only unaccompanied children exempted. Crucially, the list of countries of return, which includes countries hosting return hubs at its end, clarifies that the order of the list does not determine the sequence in which those countries may be applied. This suggests that a person may be removed to a return hub even before any attempt is made to return them to their country of origin, raising concerns as to compliance with the principle of good faith and increasing the risk of human rights violations associated with transfers to third countries.

Alternatives to return (Art. 7(9)): The Parliament retains the approach in the Commission’s proposal whereby the current possibility under Art. 6(4) of not issuing a return decision is removed from the list of situations in which such a decision is not adopted. Under the Parliament’s proposal, decisions by Member States to grant an autonomous residence permit, long-stay visa, or other authorisation granting a right to stay on compassionate, humanitarian, or other grounds may only result in the withdrawal or suspension of a return decision, rather than its non-issuance.

References to national law: Finally, the Parliament defers to national law in defining a range of measures, which sits uneasily with the nature of a regulation, as opposed to a directive, and may give rise to disproportionate and arbitrary outcomes. The measures whose applicability or scope may be determined at the national level include removal (Art. 12(1)), cooperation duties (Art. 21(2)), penalties for non-compliance with such duties (Art. 22(1)), detention (Art. 29(3)), and restrictive measures (so-called alternatives to detention) (Art. 31(2)). The notion of “security risks”, which triggers the application of multiple coercive measures, also encompasses threats defined under national law.

Concluding thoughts

The European Parliament’s report on the proposed Return Regulation reflects a markedly coercive approach towards persons in an irregular situation, raising serious concerns as to its compatibility with fundamental rights. These include the prohibition of refoulement, arbitrary detention, torture and ill-treatment, the right to due process and procedural safeguards, the rights of the child, as well as socio-economic rights, alongside broader principles of proportionality and human dignity. This approach stands in stark contrast to the Parliament’s December 2020 resolution, which reflected safeguards developed by UN human rights mechanisms and may be regarded as a counterbalance to the Commission’s 2017 recommendation on the implementation of the Return Directive. In that resolution, the Parliament emphasised that voluntary return should be prioritised over forced return; unaccompanied children should only be returned where this is demonstrably in their best interests; children should never be detained for immigration purposes; and detention must remain a measure of last resort. It also expressed concern over the widespread automatic imposition of entry bans and broad criteria for assessing the risk of absconding, and encouraged greater use of autonomous residence permits.

The current European Parliament does not appear to provide an effective counterweight to executive proposals that raise concerns from both a human rights and effectiveness perspective. While the Commission’s proposal for the Return Regulation had already been widely criticised as disproportionate and problematic from a fundamental rights standpoint, the Parliament largely retains its most contentious elements and, in many respects, further reinforces their coercive nature. In practice, the Parliament’s position is closely aligned with that of the Council, itself regarded as particularly restrictive. As a result, trilogue negotiations are expected to proceed swiftly, mirroring the accelerated adoption of recent instruments such as the regulations on safe third countries and safe countries of origin.

This marks a striking departure from the legislative dynamics during the 2005–2008 legislative process leading to the adoption of the Return Directive (discussed here, here and here). At that time, acting under the then newly established co-decision procedure, which placed it on equal footing with the Council, the European Parliament was widely regarded as a key defender of fundamental rights. It significantly strengthened safeguards in the draft Directive proposed by the Commission and, during the subsequent negotiations with the Council, played a decisive role in preventing the adoption of some of the Council’s more restrictive proposals. The institutional dynamics were famously characterised as a triad of the “Good” (Parliament), the “Bad” (Council), and the “Ugly” (Commission). Nearly two decades later, however, this configuration appears to have shifted, with all three actors now converging towards the “Bad.”

Miguel Alconero Bravo, Predoctoral Research Fellow (Formación de Personal Investigador – FPI) at the University of Valladolid*

*Part of the

Photo credit: berthgmn via wikimedia commons

Introduction

Predicating the detention of a migrant on circumstances entirely beyond his or her control seems prima facie unjust. Indeed, even a cursory reading of the provisions of EU law empowering Member States to extend the detention of a foreign national due to delays in obtaining the necessary documentation from third countries unmistakably exposes the inherent disproportionality of such a measure.

Ordering detention v. Extending detention

It must first be emphasised that both Article 15(6) of the Return Directive and Article 32(3) of the Commission proposal to overhaul EU return procedures enable Member States solely to extend a period of deprivation of liberty that has already been lawfully ordered.

Accordingly, where the conditions required under EU law for the detention of a third-country national are not satisfied in a given factual scenario, it is legally immaterial whether any of the grounds permitting a State to prolong that measure are present. If a person could not lawfully have been deprived of liberty from the outset, there is, fundamentally, no valid detention susceptible of being extended.

In this regard, Article 15(5) of the Return Directive states that “detention shall be maintained for as long a period as the conditions laid down in paragraph 1 are fulfilled and it is necessary to ensure successful removal”, whereas Article 32(1) of the proposal for a Return Regulation provides that “detention shall be maintained for as short a period as possible and for as long as the conditions laid down in Article 29 are fulfilled […]”.

Reference must therefore be made to Article 15(1) of the Return Directive, which adopts a numerus apertus approach. This provision refers in particular to two situations that may lead to the detention of a foreign national pending removal: the existence of a risk of absconding, and the possibility of identifying conduct through which the migrant seeks to avoid or hamper the preparation of return or the removal process.

Similarly, Article 29(3) of the proposed Regulation sets out, in exhaustive terms, the only situations in which the requesting State may order the detention of an irregularly staying third-country national: (a) where there is a risk of absconding determined in accordance with Article 30; (b) where the third-country national avoids or hampers the preparation of the return or the removal process; (c) where the third-country national poses security risks in accordance with Article 16; (d) in order to determine or verify his or her identity or nationality; (e) in the event of non-compliance with the measures ordered pursuant to Article 31.

Thus, the vast majority of the grounds for detention laid down in the Return Directive and in the proposed Regulation relate to scenarios in which the third-country national concerned has sought to frustrate the enforcement of return or the removal process. In essence, deprivation of liberty under both instruments is generally linked to some form of non-cooperation by the person concerned.

In this context, it must be stressed that construing the foreign national's lack of cooperation as a ground for extending detention – as Article 15(6)(a) of the Return Directive and Article 32(3) of the proposed Regulation do – makes it easier to postulate that the ground originally justifying detention still subsists. In other words, such a construction allows the authorities to establish the presence of almost any of the grounds set out in Article 15(1) of the return Directive and in Article 29(3) of the proposed Regulation for the purpose of ordering the detention of a migrant.

As a matter of fact, when a State invokes the “lack of cooperation by the third-country national concerned” in order to prolong detention, it is essentially relying on a concept whose indeterminacy allows national authorities to subsume within it many of the very grounds on which they could initially have relied to deprive that person of liberty.

In sum, extending the detention of a foreign national as a consequence of his or her lack of cooperation will, in most cases, make it possible to ascertain that the grounds for detention set out in EU law continue to be met.

Nonetheless, an exception must be made in respect of Articles 29(3)(c) and 29(3)(d) of the proposed Regulation –namely, where the person concerned is a third-country national deemed to pose a security risk, and where detention is ordered for the purpose of determining or verifying the foreign national’s identity or nationality.

In that respect, it should be noted that detention ordered on the grounds that a migrant poses a risk to public order, public security or national security bears, strictly speaking, no relation to the degree of cooperation he or she may have been willing to display during the return procedure. Irrespective of the individual’s willingness to cooperate, once a foreign national has been deprived of liberty on the basis that he or she is regarded as a threat to a State’s public order, public security or national security, detention will persist for as long as that classification remains in force.

That being so, authorities cannot invoke a migrant’s lack of cooperation to extend detention where the initial deprivation of liberty was based on security grounds or on the need to verify identity, as those circumstances do not, as such, relate to the individual’s conduct during the return procedure (unless a genuine and subsequent lack of cooperation actually materialises). Nevertheless, it is equally true that, in our view, Brussels errs in treating said situations as grounds for detention in the first place.

Against this backdrop, the pivotal question is as follows: under which of the grounds set out in either Article 15(1) of the Return Directive or Article 29(3) of the proposed Regulation could a Member State’s intention to prolong the detention of a third-country national be classified when that intention is based, purely and simply, on the likelihood that the procedure will take longer owing to delays in obtaining the required documentation from third countries?

As far as the Return Directive is concerned, the answer may lie in the numerus apertus structure of Article 15(1). Article 29(3) of the proposed Regulation, however, does not, under any circumstances, permit a third-country national to be deprived of liberty solely on the basis of delays in obtaining documents from third countries.

The conclusion follows inexorably. The wording of Article 32(3) of the proposed Regulation allowing deprivation of liberty to be prolonged “where the return procedure is likely to last longer owing to […] delays in obtaining the necessary documentation from third countries” ought to be removed. If EU law does not recognise the possibility of detaining a foreign national on that ground, it is a fortiori manifestly contrary to the principle of proportionality that such delays should then be relied upon to justify prolonging the duration of a custodial measure.

A structural contradiction

It must be acknowledged that both Article 32(3) of the proposed Return Regulation and Article 15(6)(b) of the Return Directive pursue an entirely legitimate aim: namely, to prevent, through the use of all available means, a situation in which enforcement of the return decision becomes impossible as a result of a temporary delay in receiving the necessary documentation from the requested State.

That said, it is crucial to recall that, in order for the detention of a third-country national to be prolonged, there must first exist a custodial measure resting on a valid legal basis. The analysis must therefore begin by determining whether in the factual circumstances of the case the substantive conditions laid down in EU law for depriving a migrant of liberty are actually satisfied.

Against this background, neither Article 15(1) of the Return Directive nor Article 29(3) of the proposed Regulation treats a mere delay in the completion of the administrative or diplomatic formalities necessary to obtain the required documentation from third countries as a circumstance capable of justifying an initial deprivation of liberty.

This gives rise to a stark legal contradiction: if EU law precludes the initial detention of a third-country national on the basis of administrative delays in obtaining the necessary documentation from third countries, why does Brussels permit its Member States to extend detention on that very ground? Although such delays cannot justify the initial deprivation of liberty, they may nonetheless serve to prolong the detention of a migrant.

The result is a structural paradox whereby a custodial measure that should come to an end from the moment the substantive conditions underpinning detention are no longer met is nevertheless lawfully prolonged by the requesting State under the provisions governing extension.

That tension becomes all the more striking when viewed in the light of other instruments of EU migration law. In particular, Article 11(1) of Directive 2024/1346 provides that “delays in administrative procedures that cannot be attributed to the applicant shall not justify a continuation of detention”.

For all these reasons, if the EU wished to preserve this ground for prolonging detention in the proposed Regulation, its express incorporation into Article 29(3)  as a ground, not just for extending, but also for ordering detention would be necessary.

As regards Directive 2008/115, the only interpretative solution would be to conclude that, by merely identifying the risk of absconding and the possibility that the individual may avoid or hamper the preparation of return as the two grounds that, in particular, entitle a State to deprive a migrant of liberty, that instrument implicitly leaves room for the authorities to rely on other situations as independent grounds for detention. One such additional scenario would be, precisely, delays in obtaining the necessary documentation from third countries.

However, both of those alternatives must be categorically rejected. Otherwise, the ultima ratio nature traditionally associated with the detention of an irregularly staying third-country national would be severely compromised.

In any event, the release of the individual concerned does not prevent the authorities from resorting to less restrictive measures designed to secure the enforcement of return pending receipt of the necessary documentation.

In this regard, Article 31 of the proposed Regulation explicitly provides for various alternatives to detention, which must be “proportionate to the level of the risk of absconding assessed in accordance with Article 30”. Furthermore, Article 23 of said proposal provides for a series of options which, by restricting the individual’s geographical mobility, are capable of ensuring his or her availability during the removal process.

Mahdi: When can detention be extended?

The Court of Justice of the European Union (CJEU) clarified in Mahdi the scope of the assessment that national courts must undertake before authorising any extension of a detention period on the basis of Article 15(6) of the Return Directive

In order to rule on the questions referred by the Administrativen sad Sofia-grad, the CJEU began by identifying the defining features of Article 15 of the Return Directive. It emphasised, in particular, that this provision is both “unconditional” and “sufficiently precise”, qualities which explain why its implementation by the Member States requires “no other specific elements”.

The Court then turned to the provisions whose interpretation had been requested by the referring court, namely Articles 15(3) and 15(6) of the Return Directive.

Concerning Article 15(3), the Court underlined the clarity with which its wording indicates that any detention measure prolonged over time must be subject to the “supervision” of a judicial authority. Yet the CJEU also acknowledged that the provision does not specify the precise nature of that examination, thereby making it necessary to “recall the rules deriving from Article 15 […]” in order to complete the analysis required by the question raised by the Bulgarian court.

It was at that stage that the Court brought Article 15(6) of Directive 2008/115 into the discussion.

Significantly, however, the CJEU did not confine itself to identifying the consequences flowing from the substantive conditions set out in that provision. Instead, it examined Article 15(6) in the light of Article 15(4) of the same Directive.

In so doing, the Court of Justice of the European Union introduced two decisive elements into the judicial review of the deprivation of liberty of an irregularly staying third-country national.

First, it held that, “at the time of the national court’s review of the lawfulness of

detention”, there must be “a real prospect that the removal can be carried out successfully”, having regard to the time limits laid down in Articles 15(5) and 15(6) of Directive 2008/115.

Secondly, the Court established that the authority determining whether the individual’s detention should be extended or whether he or she should be released must re-examine the substantive conditions laid down in Article 15(1) of the Return Directive. Crucially, this entails verifying that the grounds “which have formed the basis for the initial decision to detain the third-country national concerned” continue to subsist.

This point is of central importance to the present discussion, since Article 15(1) of the Return Directive makes no reference whatsoever to the possibility of detaining a foreign national solely on account of delays by third countries in issuing the necessary documentation. It follows that reliance on such a ground does not relieve the competent authorities of the burden of demonstrating, first, that there remains a “reasonable prospect of removal” and, secondly, that the substantive grounds which justified the initial deprivation of liberty continue to subsist.

The disappearance of either the reasonable prospect of removal or the substantive grounds that initially justified the detention of the third-country national necessarily renders that deprivation of liberty unlawful, as Article 15(4) of the Return Directive makes clear. That conclusion is not altered by the mere fact that the authorities continue to await the necessary documentation from other States.

The post Europe has become a strategic spectator in the Iran war first appeared on LSE European Politics.

Attila Szabó, PhD

Head of Legal Aid Service, Hungarian Civil Liberties Union

*The author assisted the lawyer representing the person concerned as an advisor in the Hungarian case analyzed in the text.  The author also used artificial intelligence to prepare the English version of the text.

Photo credit: Jorge Franganillo, via Wikimedia commons

In recent years, the Court of Justice of the European Union has increasingly engaged with trans rights. European constitutional community has followed this development. 

It appears that despite the consolidation of a GDPR interpretation aligned with trans rights, and thus with human dignity, Hungarian courts fail to understand that gender identity is not only a matter of self-determination, but also one of data accuracy. If someone presents and lives as a woman, then from the perspective of data accuracy she must also be treated as a woman, since this is the accurate data. The highest Hungarian court, however, sees this differently, and with its decision on the matter, it violates EU law.

Between 2024 and 2025, the Court of Justice of the European Union (CJEU) reshaped the landscape of trans rights in the EU through a remarkable line of cases: Mirin, Mousse, Deldits and Shipov. Taken together, these rulings reveal a structural shift in the CJEU’s approach. The Court increasingly speaks the language of “gender identity” rather than “gender reassignment,” signalling a move away from medicalised understandings of trans status and opening space for non-binary recognition. It integrates ECtHR standards as a constitutional floor while embedding trans rights across multiple doctrinal pillars: EU citizenship, free movement, privacy, equality, and data protection. What emerges is not a single breakthrough, but a coherent jurisprudential arc. One that justifies speaking of a significant doctrinal shift of trans rights in EU law.

Summary of the Hungarian Kúria’s Decision

The claimant requested the rectification of the “sex” entry in the Hungarian personal data and address register from “male” (as recorded in the civil registry at birth) to “female,” relying primarily on Article 16 GDPR (right to rectification) and explicitly invoking the CJEU’s judgment in Deldits. The claimant argued that the register in question records “sex”, not “sex at birth,” and therefore should reflect lived (social) gender identity. Since her appearance and social relations objectively correspond to a female identity, the currently recorded data were inaccurate within the meaning of Article 5 (1) d) and Article 16 of GDPR. She maintained that the data protection authority should assess whether the recorded data correspond to reality as experienced and perceived, and that EU law requires rectification even if domestic civil registry law remains unchanged.

Both the first-instance court and, on review, the Kúria (Hungarian Supreme Court) rejected the claim. (The decisions have not yet been made public; this blog post provides the first summary of them.) The decisive reasoning was structural: the personal data and address register is a secondary, derivative register whose “sex” entry is based directly on the birth registry. Under Hungarian law, the birth registry records “sex at birth,” defined biologically. Since the personal data register derives this data from the civil registry, it cannot diverge from it without undermining legal certainty and the authenticity of public registers. The Kúria held that the register does not record “lived gender identity” at all; therefore, the data cannot be considered inaccurate merely because the claimant’s current gender identity differs from the birth record. In its reading of Deldits, the GDPR right to rectification applies only to data that are inaccurate within the meaning and function of the specific register concerned. Article 16 GDPR cannot be interpreted as obliging an authority to insert new categories of data (e.g. lived gender identity) without explicit statutory authorisation, nor to assign a different substantive meaning to an existing category (“sex at birth”). The Court therefore concluded that no inaccuracy existed and that rectification was not required.

Mistakes in the decision

The conceptual distinction between “sex at birth” and “sex” undermines the exclusivity claim. The birth registry records a historical biological fact at the time of birth. By contrast, the personal data and address register functions as an operational identification database used for everyday legal and administrative interactions. Accuracy in this context serves identification and legal certainty in present-day relations. If the data recorded there do not reflect the individual’s lived and socially recognised gender, they may fail the accuracy requirement precisely because they hinder reliable identification. The fact that one dataset originates historically from another does not transform the original entry into an immutable legal truth for all future processing contexts.

Secondly, Hungarian law itself does not establish that the civil registry is the sole permissible source of “sex” data in all registers. If it would have been so it would be absolutely unchangeable. However it is not, since the statutory framework governing the personal data and address register allows updates based on legally valid rectification requests. It is fully in line with GDPR. Moreover, even Hungarian constitutional jurisprudence has recognised that legal acknowledgment of gender identity, at least outside the civil registry context, may be compatible with the Hungarian Fundamental Law. This demonstrates that the legal system does not treat the birth entry as metaphysically definitive, but as one administrative record among others.

Finally, the “hierarchy of registers” argument reverses the logic of legal certainty. Legal certainty is not preserved by maintaining inter-database consistency at the price of factual inaccuracy. Rather, certainty requires that state records correspond to verifiable social and legal reality. If necessary, consistency between registers can be achieved by differentiating between “sex at birth” (retained in the birth registry) and current “sex” or gender identity (reflected in identification databases). EU law does not require uniformity of terminology across all databases; it requires accuracy, proportionality, and effective protection of fundamental rights. Therefore, the proposition that only the birth registry may serve as the lawful source of sex-related data is neither compelled by domestic law nor compatible with the GDPR as interpreted by the Court of Justice.

The Decision in the Context of European Law

This ruling stands in notable tension with the emerging CJEU jurisprudence represented by the Deldits, Mirin, Mousse and Shipov cases. In Deldits, the CJEU held that where a register contains personal data relating to gender identity, that data must be rectified if inaccurate, and that Member States may not impose disproportionate evidentiary burdens (such as proof of surgery). The Hungarian Kúria distinguished Deldits on the basis that the asylum register there functioned as a primary identity register, whereas the Hungarian personal data register merely mirrors the civil registry’s birth-sex entry. The core of Kúria's reasoning is therefore ontological: if a register is designed to record biological sex at birth, then a divergence from lived gender identity does not render it “inaccurate.”

This is, of course, a misconception: the personal data register, which is distinct from the civil (birth) registry, exists precisely to record the data necessary for identification. In most cases, those data derive from the civil registry; however, in the case of trans persons, they do not necessarily follow from the birth registry but from factual circumstances that, through a rectification procedure, could also become officially documented facts. This state database can record birth sex data and accurate, actual data too, in parallel. Kúria argues that the registry only processes data relating to "sex" and cannot process data relating to birth gender and current gender without a change in legislation. This is true, of course, but the data controller can, in such cases, process only data relating to current gender in the registry that exists alongside the birth registry system.

Nor is Kúria's argument persuasive that it cannot order the processing of “new” data. The personal data register can, within the existing legal framework, be modified technically and administratively if in its current form it does not comply with the requirements flowing from the GDPR and EU law. The Kúria therefore did not give effect to EU law, but rather to domestic practical constraints: it effectively treated a functional system as if it were a legal norm, even though in reality such a system should adapt to legal norms, not prevent their enforcement.

From an EU law perspective, however, this formalistic register-based distinction raises deeper questions. The recent CJEU trend has emphasised substance over classification: Mirin prioritised the practical effectiveness of EU citizenship and identity coherence across registers; Mousse treated gender-related data as protected personal data subject to strict necessity and proportionality review; and Deldits framed gender identity as a legally relevant dimension of accuracy under the GDPR. Against this background, the Hungarian decision represents a restrictive reading of Article 16 GDPR, confining rectification to internal consistency within a nationally defined registry hierarchy.

In this case, we will still turn to the Hungarian Constitutional Court. However, if that body also fails to restore the possibility of effective legal enforcement under the Hungarian Fundamental Law and the binding EU law applicable on that basis, then, besides applying to the European Court of Human Rights, the only remaining option will be for the European Commission to initiate infringement proceedings and thereby compel Hungary to comply with the binding requirements of the GDPR and the Charter of Fundamental Rights. And yes, if necessary, let them add another field to the personal data register system.

Following the very recent Shipov decision, the situation is even clearer: the position of the Hungarian Supreme Court is completely untenable. In that case, the court ruled that it violates the right to free movement if a person is unable to identify themselves in another Member State with an identity document corresponding to their true gender. The current Hungarian case highlights that this violates not only the right to free movement but also the GDPR’s data accuracy rules. After all, inaccuracy is not only a problem when someone travels to another Member State. The two cases are thus based on different legal arguments, but they point to the same thing: inaccuracy causes privacy difficulties that violate the right to private life protected by the Charter of Fundamental Rights.

Dr Radosveta Vassileva, Adjunct Senior Research Fellow, UCD Sutherland School of Law

Photo credit: Raggatt2000, via wikimedia commons

On 12 February 2026, the CJEU delivered its judgment in Case C-56/25 which concerns a preliminary reference from Bulgaria raising a question about the application of the principle of primacy of EU law in view of a provision of the Procedural Rules of Bulgaria’s Constitutional Court. The CJEU judgment was rendered without an Opinion by the Advocate General, indicating that, in the eyes of the CJEU, the question was neither new, nor difficult. Moreover, the CJEU quotes its settled case law on the primacy of EU law.

However, a closer look at the Bulgarian legislation, which gave rise to the question by the Bulgarian court, especially against the broader country-specific context, reveals that the judgment seems to foreshadow inevitable conflicts between the Bulgarian and the EU legal orders in the future. From a Bulgarian perspective, the place of EU law in the hierarchy of norms is far from being conclusively determined. In parallel, the judgment demonstrates some of the flaws of a controversial constitutional reform carried out in Bulgaria in 2023, whose grand innovation was the ‘individual constitutional complaint’, which led to the hierarchy of norms dilemma of the referring court.

‘The chicken-or-the egg’ problem of the Bulgarian court

The preliminary reference was made in 2025 by the Sofia City Court, acting as a first instance in criminal proceedings, which seemingly faced a ‘chicken-or-egg’ dilemma.

The Sofia City Court was concerned that a provision of national law relevant to the qualification of the criminal offence and, respectively, the punishment in the criminal proceedings before it violated both Bulgaria’s Constitution and EU law – namely, Article 4 of Council Framework Decision 2004/757/JHA of 25 October 2004 laying down minimum provisions on the constituent elements of criminal acts and penalties in the field of illicit drug trafficking.

The Sofia City Court deemed that it was more appropriate to ask the Constitutional Court to rule on the constitutionality of the contested provision of national criminal law before it made a preliminary reference to the CJEU about the compatibility of Bulgarian criminal law with the aforementioned council framework decision. However, in its eyes, there was a catch – Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court requires that a request to it

…must contain a reasoned assessment of the applicable law, including of the consequences of the application of EU law where the contested provision or act comes within its scope.

In this context, the Sofia City Court asked the CJEU the following:

Are Article 267 TFEU, Article 94(b) of the Rules of Procedure of the Court … and the principle of the primacy of EU law … to be interpreted as meaning that, where a national court has doubts as to the compatibility of a provision of national law with EU law, and is at the same time convinced that that provision of law is [contrary to the national Constitution], it is obliged or entitled, before submitting its request for a preliminary ruling, to establish whether that provision of national law is indeed applicable in the main proceedings by making an application to the… Constitutional Court… for a declaration as to its unconstitutionality?(para 27 of judgment).

Contextual background: ‘the individual constitutional complaint’ as the grand innovation of the 2023 constitutional reform

The significance of the question by the Sofia City Court can be better appreciated against the broader context of the Bulgarian constitutional reform of 2023, which, regrettably, is neither explained in the text of the preliminary reference itself, nor in the CJEU judgment. This reform, however, made Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court, which is at the heart of the hierarchy of norms dilemma of the Sofia City Court, a legal irritant from an EU law perspective.

Before the 2023 reform, the opportunities to request a review by the Constitutional Court were ‘significantly limited and depend[ed] on the political climate in the country’ (see Radosveta Vassileva, ‘A Perfect Storm: The Extraordinary Constitutional Attack against the Istanbul Convention in Bulgaria’ (2022) 1 Osteuropa Recht 84). According to the previous wording of Article 150(1) of the Bulgarian Constitution, the Constitutional Court could only be approached by one-fifth of the Members of Parliament, the President, the Council of Ministers, the Supreme Court of Cassation, the Supreme Administrative Court, or the General Prosecutor. In some limited cases, it could be approached by the Ombudsman or the Supreme Bar Council (see the previous wording of Articles 150(3) and 150(4) of the Constitution).

To this end, one of the grand ‘innovations’ of the 2023 constitutional reform was the introduction of the ‘individual constitutional complaint’. The name given by the Bulgarian legislator to this new procedure, nevertheless, seems to be a misnomer because a private citizen can neither directly ask the Constitutional Court for constitutional review nor appeal court decisions there on constitutional grounds. The new mechanism for constitutional review, in essence, mimics the preliminary reference procedure before the CJEU. In fact, as argued in the explanatory memorandum accompanying the reform bill of 2023, ‘the situation in Bulgaria [was] paradoxical because any court (judge) could [make a preliminary reference to the CJEU], but could not turn directly to the national constitutional jurisdiction’ (p 17 of the aforementioned explanatory memorandum).

Following the 2023 reform, Article 150(2) of the Constitution stipulates:

Any court, at the request of a party to the case or on its own initiative, may refer to the Constitutional Court a request to establish an incompatibility between a law applicable to the specific case and the Constitution. The proceedings in the case shall continue, and the court, whose decision is final, hands down its judgment after the proceedings pending before the Constitutional Court have been concluded.

It is important to note that the usage of ‘may’ in the wording implies that the judge or the judicial panel in question has discretion over whether to honour such request by a party in the proceedings. In practice, it is common for courts to choose not to honour requests for preliminary references to the CJEU and/or requests to approach the Constitutional Court under the new wording of Article 150(2) coming from the parties. In other words, there are cases in which judges turn a blind eye to the likely incompatibility between Bulgarian legislation, on the one hand, and EU law and the Constitution, on the other. There are also cases in which judges can be seen in a more activist role, making references when the parties concerned did not argue any such incompatibility themselves.

The ‘ticking procedural time bomb’ left by the Bulgarian legislator

Bulgaria’s Constitution, adopted in 1991, clearly specifies the hierarchy of norms. Its Article 5(1) explicitly states that it is the ‘supreme law’, while its Article 5(3) stipulates:

International treaties, ratified in accordance with the constitutional procedure, promulgated, and having entered into force for the Republic of Bulgaria, are part of the country’s domestic law. They take precedence over those provisions of domestic legislation that contradict them.

Hence, the Constitution takes precedence over any international treaty and Bulgarian law while international treaties take precedence over Bulgarian laws that contradict them. Bulgaria has, of course, ratified its EU Accession Treaty of 2005.

Furthermore, the status and place of CJEU case law in the hierarchy of norms is subject to debate from a national perspective. CJEU’s case law is not defined as binding in Bulgarian legislation, unlike the case law of the Constitutional Court or the decrees and decisions on interpretation by the country’s supreme courts. However, the Code of Civil Procedure allows cassation on the grounds of violation of the case law of the CJEU (see Article 280(1), point 2).

It is notable that throughout the years, Bulgaria’s Constitutional Court has tried to avoid direct confrontation with the CJEU. In its own case law, however, it has argued that it is up to each entity requesting constitutional review to determine the applicable law, including EU law, to the proceedings before it. It has also stressed that the determination of the applicable law should always precede requests for constitutional review (see, for instance, the detailed reasoning in Ruling 2 of 24 February 2022 by the Constitutional Court).

In this light, the ‘individual constitutional complaint’ can be seen as a ‘ticking procedural time bomb’ since the reform of 2023 left Article 5 of the Constitution untouched, while expanding the jurisdiction of the Constitutional Court itself, which can now receive requests directly from judges. In practice, this ‘individual complaint’ means a chance for more frequent encounters and, respectively, potential clashes between Bulgarian constitutional law and EU law in ordinary Bulgarian courts.

The legislative choice not to tamper with Article 5, in turn, was informed by the restrictions on amendments imposed by the Constitution itself – any amendments to Article 5 require the convocation of a ‘grand national assembly’ as opposed to a ‘regular national assembly’ (on the difference between them, see here; see also Article 158 of the Constitution). Grand national assemblies are notoriously difficult to convene. The Constitution of 1991 was adopted by such an assembly.

CJEU’s judgment as an exercise in judicial diplomacy

In the judgment in Case C-56/25, the CJEU held that Article 267 TFEU, the principle of the primacy of EU law and Article 94(b) of the Rules of Procedure of the Court of Justice do not preclude procedural rules, such as Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court (para 61 of judgment). Leading authorities on the rule of law like Professor Laurent Pech have already called the judgment ‘unusually generous’, especially in view of challenges to CJEU’s jurisdiction from other EU member states, such as Poland.

Yet, when one reads the reasoning of the judgment in light of the particularities of the Bulgarian constitutional order, it seems that the ‘generosity’ came with a few price tags – namely, demands for a voluntary surrender of jurisdiction by the Bulgarian Constitutional Court. These requests, nevertheless, may not bear the fruit intended by the CJEU, precisely because they seem oblivious of context, including national legislation.

One of the reasons why the CJEU was not troubled by the requirements of Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court was the Constitutional Court’s alleged obligation to make preliminary references itself. Namely, the CJEU stated: ‘…if a constitutional court is seised of a request for a review of the constitutionality of a provision of national law coming within the scope of EU law, that court is in principle obliged to make a reference to the Court of Justice for a preliminary ruling, in accordance with the third paragraph of Article 267 TFEU…’ (para 57 of judgment). Bulgaria’s Constitutional Court, however, has no record of making preliminary references to the CJEU and it is highly unlikely that it will endeavour to make such references in the future if an amendment to Article 5 of the Constitution is not made. That is because the Constitutional Court sees its own role solely as the guardian of the Bulgarian Constitution. First, the Constitutional Court has the monopoly on authoritative, binding interpretation of the Constitution (Article 149(1), point 1 of the Constitution). Second, in case of conflict between the Constitution and EU law, the Constitution prevails according to its own text. Third, as highlighted above, the Constitutional Court has stated in its own settled case law that the establishment of the applicable EU law is in the prerogative of the court examining the dispute on the merits. On the one hand, this approach protects the Constitutional Court’s jurisdiction. On the other hand, it prevents direct confrontation with the CJEU – it is up to the ordinary courts to identify and raise any questions about the compatibility between EU law and constitutional law and to consider how such conflict affects the outcome of the case.

It is also interesting that the CJEU reminds that a ‘national court is required, in order to ensure the full effectiveness of the rules of EU law, to disregard, in the dispute before it, the rulings of a national constitutional court which refuses to give effect to a judgment given by way of a preliminary ruling by the Court of Justice…’ (para 59 of judgment). From an EU perspective, this approach has numerous merits, such as empowering judges in lower courts to disregard controversial decisions by constitutional courts, especially when they are unlawfully composed. Nevertheless, in the Bulgarian context, beyond the invitation to disregard key express provisions of the Bulgarian Constitution, this conclusion ignores important practical aspects. For example, pursuant to Article 280(1), point 2 of the Bulgarian Code of Civil Procedure, non-compliance with case law by the Constitutional Court serves as grounds for cassation. While the Code of Criminal Procedure does not contain such explicit reference to constitutional case law, its Article 348 allows cassation for ‘violations of the law’ broadly conceived.

A thorny path ahead for the primacy of EU law

Overall, the path ahead for the primacy of EU law in Bulgaria appears rather thorny. CJEU’s judgment in Case C-56/25 may be seen as a precursor to a series of legal and political challenges. It foreshadows inevitable clashes between EU law and Bulgarian constitutional law in ordinary courts, which are facilitated by the shortcomings of the 2023 constitutional reform.

It is also doubtful to what extent the approach, which the CJEU endorses in the judgment, will contribute to affirming the primacy of EU law in Bulgaria. Ensuring such primacy on paper requires important constitutional amendments by a grand national assembly as well as a large-scale legislative reform. It also necessitates a change of mentalité in practice on behalf of both Bulgarian judges and the CJEU itself.

First, CJEU case law remains terra incognita for many jurists because of their educational and professional background. Moreover, for more than a decade following Bulgaria’s EU accession in 2007, preliminary references from the country came from a very small circle of judges (see Aleksander Kornezov, ‘Ten years of preliminary references – a critical review and appraisal’(2017) Evropeiski praven pregled).

Second, between 2018-2022, a vast number of preliminary references were made by Bulgaria’s specialised criminal court, which had all features of an extraordinary repressive tribunal, and used this EU law mechanism primarily as a tool to affirm its legitimacy in light of criticism of its abusive practices. In 2022, this court was dissolved for good for undermining the rule of law upon the proposal by a short-lived opposition government and after years of public discontent. As a result, doubts about the status of its case law and the EU judgments resulting from its references, which are intertwined with it, are on the rise. So is the bitter feeling that the CJEU empowered ‘non-judges’ the way it did vis-à-vis Poland (see here).

Third, the CJEU has acted as a Pontius Pilate, washing its hands of responsibility, on key matters pertaining to Bulgaria’s rule of law backsliding and human rights abuses. It has a record of prioritising formalism and refusing to address the issues at their core, thus discouraging judges from raising politically sensitive questions (on judicial independence, see here; on standards of proof in pretrial detention, see here).

To what use is then primacy for primacy’s sake? This is a question which many judges from Bulgaria’s ordinary courts may ask in the aftermath of CJEU’s judgment in Case C-56/25.

The author would like to thank Prof Laurent Pech for his helpful comments and suggestions on an earlier draft.

The post In the age of AI, citizens may become more critical of human decision-makers first appeared on LSE European Politics.

Mariana Lima Rodrigues Carneiro (Masters in European Union Law from the School of Law of University of Minho)

The deployment of xAI’s Grok chatbot has become a focal point of systemic risk within the European digital landscape. The European Commission first opened formal proceedings against X in December 2023.[1] In January of 2026, the scope of this regulatory oversight was significantly expanded under the Digital Services Act (DSA) to investigate Grok’s functionalities.[2] This investigation specifically targets risks such as the dissemination of non-consensual sexual deepfakes and antisemitic discourse. These controversies reveal a programmed tendency towards neutral language that masks structural biases within AI systems.[3]

This article explores how this systemic opacity creates an insurmountable barrier for individuals seeking legal redress against algorithmic discrimination. The core objective is to analyse the failure of the current reversal of the burden of proof mechanism, as provided by European anti-discrimination directives, when faced with high-dimensional mathematical optimisation. Ultimately, this text examines the necessity of technical solutions to harmonise automated processing with the values of justice and equality that underpin the European legal order.

1. The “black box effect”

Artificial Intelligence (AI) is divided into several subareas, among which Machine Learning (ML) represents a major one. ML focuses on developing algorithmic models capable of making predictions or decisions based on data sets. Deep Learning (DL) constitutes a specialized subset of ML based on a conceptual model of the human brain, utilizing artificial neural networks structured into multiple hidden layers, commonly referred to as Deep Neural Networks (DNNs).[4] Consequently, all DL systems are ML systems, but not all ML systems are DL systems.

The predictive process begins with an input layer, which corresponds to the raw information entering the system. This is followed by numerous hidden layers where each “neuron” encodes a mathematical function that transforms the input into an output, creating a final prediction. The term “deep” refers to the high number of these hidden layers. Crucially, the more layers a system has, the more complex and obscure the decision-making process becomes, making it difficult for humans to understand how the final result was reached.[5]

This interaction results in the “black box effect”, characterised by the lack of transparency and interpretability regarding the system’s internal processes. While present in ML, this effect is exacerbated in DL systems due to the high complexity of the DNNs. As these algorithms are increasingly used in high-risk areas, such as conducting recruitment or assessing credit eligibility, the lack of transparency may violate fundamental legal principles, including due process and equality before the law.[6]

Ultimately, there is a fundamental trade-off between precision and explainability. High-dimensional mathematical optimisation often transcends human cognitive capacity, creating an incompatibility between the system’s efficiency and human semantic interpretation.[7] As systems evolve and acquire new data, their layers and connections deepen, further intensifying their structural opacity.[8] Therefore, as accuracy increases, transparency typically decreases, leaving individuals in a state of legal defencelessness when faced with automated decisions.[9]

2. Case studies: AI biases in practice

Algorithmic systems are often implemented under the promise of efficiency, precision, and neutrality. However, despite this aura of impartiality, these systems can internalise and reproduce the prejudices and inequalities present in society. Training data is not neutral; it inevitably reflects structural asymmetries and biases. Without corrective mechanisms, algorithms tend to amplify these patterns, perpetuating discrimination against historically vulnerable groups.[10]

Biases may emerge at any stage of development, including data labelling[11] or attribute selection.[12] Even when datasets seem impartial, they can mask a history of discrimination. Ultimately, automation cloaks these decisions in an appearance of technical neutrality, making their scrutiny and legal challenge far more difficult. This opacity hinders the establishment of a clear causal link between the software functioning and unfavourable treatment, effectively rendering the right to non-discrimination unenforceable in practice.[13]

To comprehend the specific ways in which these distortions manifest, it is essential to apply the taxonomy established by Batya Friedman and Helen Nissenbaum, who categorised algorithmic bias into three distinct types: pre-existing, technical, and emergent.[14]

Pre-existing bias has its roots in social institutions, practices, and attitudes that exist prior to the creation of the system, effectively acting as a digital mirror of historical prejudices. A clear example of this is the COMPAS system used in US courts, which demonstrated a tendency to incorrectly flag Black defendants as high-risk compared to White defendants, since the training data reflected racial discriminatory patterns in the criminal justice system.[15] Technical bias arises from flaws in the design stage or the lack of representativeness in training databases. This was evident in Amazon’s discontinued recruitment tool, which penalised resumes containing the word “woman” because it was trained on historical data dominated by male applicants.[16] Finally, emergent bias occurs during real-world use as a result of interaction between the system and its users. The Tay chatbot remains a paradigm of this category, as it rapidly assimilated and reproduced offensive messages from users who manipulated its learning mechanism in real time.[17]

Building on these manifestations, the impact of algorithmic bias is further evidenced in specific industry sectors. The 2019 Apple Card case highlighted the dangers of systemic opacity in credit access. Users reported that the black box algorithm granted men significantly higher credit limits than their wives, even when they shared bank accounts. Representatives from Apple and Goldman Sachs were unable to explain the decision-making process because the model’s complexity hindered its interpretability.[18] While a subsequent investigation by the NYSDFS concluded that gender was not a direct variable, the case still raised ethical concerns.[19]

Similarly, the Optum Algorithm used in US hospitals revealed how a monetary bias can convert into a racial one. The system used healthcare costs as a proxy for health needs. Since White patients historically had higher medical expenses due to structural socioeconomic privileges, the algorithm incorrectly identified them as being “sicker” than Black patients with similar clinical conditions, who had less access to services. Consequently, only 18% of Black patients were selected for intensive care programs when the actual need was 47%.[20]

Another significant area of concern involves the use of facial recognition systems, since they present some of the most critical risks to fundamental rights. These algorithms suffer from a lack of diversity in their databases, leading to significantly higher error rates for people with darker skin tones, particularly women.[21]

A test conducted by the ACLU using Amazon’s Rekognition algorithm found that the software incorrectly matched 28 members of the US Congress with a mugshot database. Nearly 40% of these false matches involved people of colour, despite them making up only 20% of Congress.[22] Similar technical failures have already produced real-life consequences. In a well-known case, Robert Williams, a Black man, was wrongfully arrested in Detroit and held for 30 hours after a facial recognition system incorrectly identified his driver’s license photo as that of the shoplifter. These cases underscore the dangers of blind trust in automated identification systems, especially in the context of criminal justice.[23]

3. Algorithmic discrimination and the burden of proof

The principle of non-discrimination is a pillar of the European Union legal order, operationalised through a comprehensive framework that includes Directive 2000/43/EC, Directive 2000/78/EC, Directive 2004/113/EC, and Directive 2006/54/EC. These instruments aim to ensure equality of treatment by prohibiting both direct and indirect discrimination across various domains, from employment to the provision of goods and services.

Direct discrimination occurs when a person of a protected group is treated less favourably than another in a comparable situation based on a protected characteristic. In contrast, indirect discrimination arises when an apparently neutral provision or practice puts a person of a protected group at a particular disadvantage.[24] Within the digital ecosystem, algorithmic discrimination is predominantly indirect, as systems rarely rely on explicit protected traits but instead use neutral data points that function as proxies for those characteristics.

The structural opacity of the black box creates significant challenges for the traditional application of these legal concepts. When a decision is the result of high-dimensional correlations rather than human intentionality, identifying the discriminatory factor becomes an exercise in technical forensic analysis. This complexity leads to an acute information asymmetry between the technology provider and the individual, where the victim is often unable to demonstrate that they have been subject to a discriminatory outcome. This lack of transparency effectively hinders the ability to establish the necessary facts to initiate litigation.[25]

To ensure the right to effective judicial protection, EU law provides for the reversal of the burden of proof. Once a plaintiff establishes facts from which it may be presumed that there has been discrimination, it is for the respondent to prove that there has been no breach of the principal of non-discrimination.[26] However, in the context of AI use, this mechanism faces a probatio diabolica. The plaintiff struggles to provide even the initial prima facie evidence because they cannot access or interpret the algorithmic logic. This creates an impediment to access to justice where the burden of proof remains an insurmountable barrier.

Furthermore, the respondent also faces emerging challenges. Even when acting in good faith, a defendant may struggle to prove the absence of bias due to the systems complexity and its opaque nature. This creates a dual burden: while the plaintiff is trapped by the inability to see inside the black box, the defendant is often unable to provide a human-understandable explanation for the automated output.[27] Consequently, the effectiveness of the right to non-discrimination depends on evolving current evidentiary standards to account for the unique technical realities of AI.

4. The European legal architecture

The European Union has developed a multifaceted regulatory strategy to tackle some of the challenges of algorithmic bias. The General Data Protection Regulation (Regulation 2016/679) establishes the foundational layer, specifically through Article 22, which provides individuals with the right not to be subject to decisions based solely on automated processing that produce legal or significantly similar effects.[28] This provision ensures a baseline of human intervention and the right to contest automated outcomes. Although the right to an explanation is not explicitly named in Article 22, a systemic interpretation of Articles 13, 14, and 15 in conjunction with Article 22 effectively establishes such a right. This interpretation ensures that data subjects receive meaningful information about the logic involved in automated decisions, acting as a tool to bridge the knowledge gap and providing the individual with the transparency needed to scrutinize the decision.[29]

The AI Act (Regulation 2024/1689) represents the most significant advancement in this field. By adopting a risk-based classification, it mandates rigorous compliance for high-risk systems, particularly those used in employment and essential services. Article 11 requires technical documentation and Article 13 requires transparency, while Article 86 creates a specific right to an explanation for individuals.[30] This right is essential to mitigate the information asymmetry between the technology provider and the person affected, allowing for a better understanding of the logic behind the automated decision.

Regarding liability, the Product Liability Directive (Directive 2024/2853) has been modernised to include software, ensuring that discriminatory algorithmic outputs can be classified as defective products.[31] Alongside this, the Proposal for an AI Liability Directive sought to introduce rebuttable presumptions of causality to ease the burden of proof.[32] However, the 2025 Commission Work Programme confirmed the abandonment of this specific legislative path, marking a significant setback in the creation of a specialised liability regime for AI.[33] Despite this, the Representative Actions Directive (Directive 2020/1828) empowers consumer organisations to challenge systemic biases, moving the defence of non-discrimination from the individual to the collective sphere.[34]

5. Technical solutions for algorithmic fairness: The CDD metric

The inherent opacity of black box systems constitutes a significant technical and legal obstacle to the reversal of the burden of proof, a cornerstone of EU anti-discrimination law. In response to this probatory impasse, data science has developed various fairness metrics aimed at mitigating bias and ensuring equitable decisions. However, a conceptual gap remains between the contextual justice required by European jurisprudence and the rigid mathematical rules of technical metrics.[35]

Most traditional metrics, such as Predictive Parity or Equalized Odds,[36] often rely on a logic of formal equality. As Sandra Wachter, Brent Mittelstadt, and Chris Russell argue, these can be classified as bias preserving metrics because they often perpetuate the status quo by ignoring historical and structural disadvantages present in the training data. In contrast, bias transforming metrics align more closely with the EU concept of substantive equality by actively seeking to correct social inequalities.[37]

In this context, the Conditional Demographic Disparity (CDD) metric emerges as an insightful tool that warrants further exploration. This metric works by identifying residual statistical differences between social groups only after accounting for relevant, legitimate and determining variables.[38] By doing so, it evaluates whether individuals in similar conditions of merit still experience systematic disparities. This approach aligns with the gold standard established by the CJEU in the Seymour-Smith case, which demands a comprehensive comparison to evaluate discriminatory impacts.[39]

By producing objective summary statistics, CDD enables claimants to gather the necessary evidence to establish a prima facie case, effectively overcoming the black box effect.[40] Furthermore, the metric functions as a preventive instrument, allowing developers to detect and mitigate biases during the design phase.

Ultimately, the implementation of such metrics depends on a socio-technical approach that seeks to integrate technological innovation with legal, social, and ethical principles.[41] This synergy allows for a more robust protection of fundamental rights, ensuring that algorithmic processing is harmonised with the values of justice and equality that underpin the European legal order.

6. Final considerations

The increasing automation of high-impact decisions risks rendering the reversal of the burden of proof ineffective. The black box effect and information asymmetry create an almost insurmountable evidentiary barrier, making the task of establishing a prima facie case nearly impossible. While the AI Act provides a right to an explanation, this remains informative rather than evidentiary. The abandonment of the AI Liability Directive leaves a significant gap, as the Product Liability Directive and Representative Actions Directive were not specifically designed for the complexity of AI. To address these shortcomings, a socio-technical approach is required to operationalise evidentiary means.

Cases such as Albania’s digital minister, Diella, reveal the fallacy of technical neutrality.[42] We must reject the notion that algorithms are purely objective and impartial tools, recognising that they can reproduce or even amplify pre-existing human biases.[43] As Cathy O’Neil warns, the “privileged, we’ll see time and time again, are processed more by people, the masses by machines.”[44] Ultimately, the safeguard of fundamental rights depends on demanding from algorithms the same scrutiny and responsibility expected from human decisions.

[1] European Commission, “Commission opens formal proceedings against X under the Digital Services Act”, Brussels, 18 December 2023, https://ec.europa.eu/commission/presscorner/detail/en/ip_23_6709.

[2] European Commission, “Commission investigates Grok and X’s recommender systems under the Digital Services Act”, Brussels, 26 January 2026, https://ec.europa.eu/commission/presscorner/detail/en/ip_26_203.

[3] Janga Bussaja, “Analyzing Grok 4’s engagement with racism: a case study in AI fragility and deception”, 11 July 2025, http://dx.doi.org/10.2139/ssrn.5348379.

[4] Maciej Satkiewicz, “Towards white box deep learning”, 2024. Warsaw: [s.n.] (Report no. 2403.09863), https://arxiv.org/pdf/2403.09863.

[5] Fabio Gagliardi Cozman & Dora Kaufman, “Viés no aprendizado de máquina em sistemas de inteligência artificial: a diversidade de origens e os caminhos de mitigação”, Revista USP. São Paulo, no. 135 (2022): 195-210.

[6] Mônia Clarissa Hennig Leal & Lucas Moreschi Paulo, “Algorítmos discriminatórios e jurisdição constitucional: os riscos jurídicos e sociais do impacto dos vieses nas plataformas de inteligência artificial de amplo acesso”, Revista de Direitos e Garantias Fundamentais. ISSN 2175-6058, vol. 24, no. 3 (2023), doi: 10.18759/rdgf.v24i3.2311.

[7] Dora Kaufman, Tainá Junquilho, Priscila Reis, “Externalidades negativas da inteligência artificial: conflitos entre limites da técnica e direitos humanos”, Revista de Direitos e Garantias Fundamentais. ISSN 2175-6058, vol. 24, no. 3 (2023), doi: 10.18759/rdgf.v24i3.2198.

[8] Dora Kaufman, Tainá Junquilho, Priscila Reis, “Externalidades negativas da inteligência artificial…”, 52.

[9] Alessandra Silveira, “Automated individual decision-making and profiling [on case C-634/21 – SCHUFA (Scoring)]”, UNIO – EU Law Journal, vol. 8, no. 2 (2023): 74–85, https://doi.org/10.21814/unio.8.2.4. 

[10] Sandra Wachter, Brent Mittelstadt, Chris Russell, “Bias preservation in machine learning: the legality of fairness metrics under EU non-discrimination law”, West Virginia Law Review, vol. 123, no. 3 (2021), https://researchrepository.wvu.edu/wvlr/vol123/iss3/4/.

[11] Fabio Gagliardi Cozman & Dora Kaufman, “Viés no aprendizado…”.

[12] Jeffrey Dastin, “Insight – Amazon scraps secret AI recruiting tool that showed bias against women”, Reuters, 11 October 2018, https://www.reuters.com/article/us-amazon-com-jobs-automation-insight/amazon-scraps-secret-ai-recruiting-tool-that-showed-bias-against-women-idUSKCN1MK08G/.

[13] Sandra Wachter, Brent Mittelstadt, Chris Russell, “Why fairness cannot be automated: bridging the gap between EU non-discrimination law and AI”, Computer Law & Security Review. ISSN 0267-3649, vol. 41 (2021): 105567, http://dx.doi.org/10.2139/ssrn.3547922.

[14] Batya Friedman, Helen Nissenbaum, “Bias in computer systems”, ACM Transactions on Information Systems, vol. 14, no. 3 (1996): 330–347, http://doi.org/10.4324/9781315259697-23.

[15] Jeff Larson, Surya Mattu, Lauren Kirchner and Julia Angwin, “How we analyzed the COMPAS recidivism algorithm”, ProPublica, 23 May 2016, https://www.propublica.org/article/how-we-analyzed-the-compas-recidivism-algorithm.

[16] Jeffrey Dastin, “Insight – Amazon scraps…”.

[17] Dave Lee, “Tay: Microsoft issues apology over racist chatbot fiasco”, BBC News, 25 March 2016, https://www.bbc.com/news/technology-35902104.

[18] Chris Wiltz, “The Apple card is the most high-profile case of AI bias yet”, Design News, 13 November 2019, https://www.designnews.com/artificial-intelligence/the-apple-card-is-the-most-high-profile-case-of-ai-bias-yet. 

[19] Ian Carlos Campbell, “The Apple card doesn’t actually discriminate against women, investigators say”, The Verge, 24 March 2021, https://www.theverge.com/2021/3/23/22347127/goldman-sachs-apple-card-no-gender-discrimination.

[20] Ziad Obermeyer, et al., “Dissecting racial bias in an algorithm used to manage the health of populations”, Science. ISSN 0036-8075, vol. 366, no. 464 (2019), Doi.org/10.1126/science.aax2342.

[21] Patrick Grother, Mei Ngan, Kayee Hanaoka, “Face recognition vendor test part 3: demographic effects”. NISTIR 8280, December 2019, https://doi.org/10.6028/NIST.IR.8280.

[22] Jacob Snow, “Amazon’s face recognition falsely matched 28 members of Congress with mugshots”, ACLU News & Commentary, 26 July 2018, https://www.aclu.org/news/privacy-technology/amazons-face-recognition-falsely-matched-28.

[23] Kashmir Hill, “Facial recognition led to wrongful arrests. So Detroit is making changes”, The New York Times, 29 June 2024, https://www.nytimes.com/2024/06/29/technology/detroit-facial-recognition-false-arrests.html.

[24] European Union Agency for Fundamental Rights (FRA), “Handbook on European non-discrimination law”, 21 March 2018, https://fra.europa.eu/en/publication/2018/handbook-european-non-discrimination-law-2018-edition.

[25] Sandra Wachter, Brent Mittelstadt, Chris Russell, “Why fairness…”.

[26] Council Directive 2000/43/EC of 29 June 2000 implementing the principle of equal treatment between persons irrespective of racial or ethnic origin.

[27] Alessandra Silveira, “Automated individual…”.

[28] Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data (General Data Protection Regulation).

[29] Tiago Sérgio Cabral, “AI and the right to explanation: three legal bases under the GDPR”, in Data protection and privacy: enforcing rights in a changing world, ed. Dara Hallinan, Ronald Leenes and Paul De Hert (Hart Publishing, 2021), 29-56, https://doi.org/10.5040/9781509941780.ch-002.

[30] Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act).

[31] Directive (EU) 2024/2853 of the European Parliament and of the Council of 23 October 2024 on liability for defective products.

[32] Proposal for a Directive of the European Parliament and of the Council on adapting non-contractual civil liability rules to artificial intelligence (AI Liability Directive).

[33] Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, Commission work programme 2025 – Moving forward together: a bolder, simpler, faster Union, Strasbourg, 11.2.2025, COM(2025) 45 final, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52025DC0045&qid=1761126737792.

[34] Directive (EU) 2020/1828 of the European Parliament and of the Council of 25 November 2020 on representative actions for the protection of the collective interests of consumers.

[35] Sandra Wachter, Brent Mittelstadt, Chris Russell, “Why fairness…”.

[36] Alessandro Castelnovo, et al., A clarification of the nuances in the fairness metrics landscape. Scientific Reports, vol. 12, no. 1 (2022), doi: 10.1038/s41598-022-07939-1.

[37] Sandra Wachter, Brent Mittelstadt, Chris Russell, “Bias preservation…”.

[38] Sandra Wachter, Brent Mittelstadt, Chris Russell, “Why fairness…”.

[39] Judgment of the Court Seymour-Smith and Perez, 9 February 1999, case C-167/97, ECLI:EU:C:1999:60.

[40] Sandra Wachter, Brent Mittelstadt, Chris Russell, “Why fairness…”.

[41] Reva Schwartz, et al., “Towards a standard for identifying and managing bias in artificial intelligence”, Special Publication (NIST SP), National Institute of Standards and Technology, Gaithersburg, MD, [online], https://doi.org/10.6028/NIST.SP.1270.

[42] Pascale Davies, “Albania appoints world’s first AI government ‘minister’ to root out corruption”, Euronews, 12 September 2025, https://www.euronews.com/next/2025/09/12/albania-appoints-worlds-first-ai-government-minister-to-root-out-corruption.

[43] Mônia Clarissa Hennig Leal & Lucas Moreschi Paulo, “Algorítmos discriminatórios e jurisdição constitucional…”.

[44] Cathy O’Neil, Weapons of math destruction: how big data increases inequality and threatens democracy (New York: Broadway Books, 2017), 6.

Picture credit: by UMA media on pexels.com.

The post Are Russia and China losing their grip on the Middle East? first appeared on LSE European Politics.

The post How should governments respond to the Iran energy crisis? first appeared on LSE European Politics.

The post The rise and fall of Iran’s forward defence strategy first appeared on LSE European Politics.

The post Jürgen Habermas and the case for a supranational democracy in Europe first appeared on LSE European Politics.

Georgios Athanasiou, PhD Researcher, University of Antwerp

Photo credit: Julian Lupyan, via Wikimedia Commons

The Chios migrant boat shipwreck of 3 February 2026 exemplifies the acute tensions between Greece’s increasingly securitized border management and its obligations under EU law, the ECHR, and international maritime conventions such as the 1979 Search and Rescue (SAR) Convention. More specifically, off the coast of Chios island, a Hellenic Coast Guard patrol vessel collided with an inflatable boat carrying approximately 39 Afghan nationals, resulting in 15 deaths and 24 injuries, including 11 minors, and cases of miscarriage. All of the deaths were attributed to severe head trauma rather than drowning, per up-to-date autopsy reports, with survivors claim that the coast guard did not offer any prior warning or communication before ramming the migrant boat, contrary to official claims of the migrants’ speedboat initiating contact. Interestingly, the on boat cameras of the patrol vessel had been deactivated.   

This event parallels the 14 June 2023 Pylos shipwreck, Europe’s deadliest maritime migration tragedy, where over 500 lives (mainly Syrian, Pakistani, and Egyptian) were lost after the overcrowded trawler Adriana capsized, allegedly due to Coast Guard towing maneuvers following delayed rescue operation, despite prior distress alerts. The ongoing criminal proceedings in Greece have charged 17 Coast Guard personnel, including the rescue vessel captain, with felony offenses such as endangering lives and contributing to the shipwreck,. This development appears to be part of a systematic attempt to portray the eastern Mediterranean migration route as inherently life-threatening for asylum seekers, thus reflecting a pervasive securitization narrative guiding border policies of the Greek government that overshadows State accountability.

Legal Parallels and Accountability Gaps

From a legal aspect, both incidents implicate Greece’s positive obligations under Article 2 ECHR (right to life), requiring States to safeguard lives within their jurisdiction, including during maritime interceptions, and conduct effective, independent investigations into fatalities. The ECtHR has repeatedly held Greece accountable in analogous cases: in Safi and Others v. Greece, for inadequate protection and probing of a sunk migrant boat; Alkhatib and Others v. Greece, for excessive lethal force lacking “absolute necessity” and deficient regulatory frameworks for Coast Guard firearms use. Such repeated failures in border management operations seem to formulate a consistent pattern of action of the Greek authorities in handling migrant routes, in an attempt to not allow migrants to enter Greek territory/territorial waters. In this sense, although Article 3 ECHR (prohibition of inhuman/degrading treatment) further prohibits collective expulsions or pushbacks, this practice has been deemed systematic by Greek authorities in A.R.E. v. Greece (also see, here).

Under EU law, the Qualification and Asylum Procedures Directives respectively mandate the upholding of the principle of non-refoulement and individual assessments of asylum applications, while Article 4 of Protocol 4 of the ECHR and Article 19 CFR bar collective expulsions. Meanwhile, the SOLAS and SAR Conventions impose duties to render assistance “without delay” to persons in distress, disembarking them to a place of safety, irrespective of nationality or the existence of a right to enter the country. Hence, any form of interception framed as SAR mission cannot justify pushbacks or endangering the lives of migrants.

Greece’s Restrictive Policies and Criminalization of Solidarity

Domestically, this incident aligns with broader migration policy tendencies, as Greece has instrumentalized criminal law in an attempt to restrict migration, rendering irregular entry, stay and exit of the country a felony punishable with up to 5 years of imprisonment coupled with a minimum fine of €5,000 (Law 5226/2025 Government Gazette Α' 154/8.9.2025). Similarly, rejected asylum seekers face administrative fines up to €10,000, as well as up to five-year sentences or electronic ankle monitoring. Hence, the 2025 deportation law, hailed as Europe’s most stringent, essentially attempts to streamline expulsions of “economic migrants,” given that long-term regularization after 7 years of stay in the country is equally abolished.

This framework cannot be dissociated from Greece’s post-2019 migration hardening: escalated border fortifications (Evros 35 km wall), freezing of asylum applications, and systematic pushbacks exceeding 540 incidents between 2020-2022 (also see, here). Hence, high-seas shipwrecks, such as the Chios and Pylos lethal incidents, epitomize how this apparatus practically overrides positive obligations under the ECHR, as well as international humanitarian and maritime law, subordinating the protection of life at sea to national security imperatives.

The Greek Government defends its approach as prevention of illegal entry, invoking safe third country safeguards, especially for migrants arriving from Turkey, yet these yield no derogation from non-refoulement or collective expulsion bans. Hence, the Government’s approach in migration policies embodies a “fortress mentality,” which, coupled with its recent attempts to criminalize solidarity, further sets in danger the lives of migrants attempting to cross the Eastern Mediterranean route. A prominent example of this criminalization tendency include the recent Lesbos case against 24 rescuers, who were acquitted after years on charges, like espionage and smuggling, that carried up to 20-year sentences. Similarly, Norwegian activist Tommy Olsen faced an arrest warrant in February 2026 for documenting pushbacks via Aegean Boat Report, accused of criminal organization. Finally, a February 2026 migration law amendment (Law 5275/2026, Government Gazette Α’ 17/06-02-2026) makes NGO membership an aggravating factor, escalating misdemeanors (e.g., facilitation of stay) to felonies with fines exceeding €100,000, constituting the largest criminalization of solidarity in the EU.

Analysis

It is apparent that the Chios tragedy, when assessed alongside the Pylos shipwreck, does not constitute an isolated operational failure but rather indicative of a structural recalibration of border governance in the Eastern Mediterranean. The shift from the enforcement of search-and-rescue obligations to human rights violations at the EU’s external borders under SAR cover reveals a normative inversion: life-saving obligations are operationalized through a security prism that treats irregular entry primarily as a threat vector rather than a protection trigger.

At the doctrinal level, Article 2 ECHR imposes both substantive and procedural duties on States. More specifically, from a substantive aspect, States shall refrain from unlawful deprivation of life and adopt preventive operational measures where authorities knew or ought to have known of a real and immediate risk. Meanwhile, procedural obligations mandate to conduct prompt, effective, and independent investigations capable of leading to accountability. In both Chios and Pylos, the central legal question is whether Greek authorities fulfilled the due diligence threshold required during maritime interception. The reported deactivation of onboard cameras in Chios and the delayed rescue response in Pylos indicate the State’s unwillingness to comply with these operational standards.

In a similar vein, the 2026 legislative reform represents an internal consolidation of the securitization paradigm. By reclassifying irregular entry and facilitation-related conduct as felonies and elevating NGO affiliation to an aggravating factor, the Greek legal framework operationalizes criminal law as a migration-management instrument, fully adopting a “Crimmigration” approach, that is the convergence of criminal and immigration enforcement logics, as border management framework.

In this sense, the prosecution of humanitarian actors in the Lesbos case and proceedings against figures associated with monitoring networks reinforce a chilling effect on civil society oversight. When accountability mechanisms (NGO monitoring, documentation of pushbacks) are suppressed, the evidentiary architecture for fundamental rights protection is simultaneously weakened. In practical terms, criminalization of solidarity indirectly facilitates impunity.

Greece’s approach cannot be decoupled from the broader EU externalization strategy. Financial and operational support through Frontex, coupled with political endorsement of deterrence metrics (reduced arrivals as “success indicators”), generates structural incentives that privilege interdiction over protection. In this context, it appears that the Eastern Mediterranean has become a testing ground for this hybrid governance model of the EU’s external borders. This primarily includes operational opacity (restricted access, disabled recording systems), normative elasticity (expansive security justifications), and penal reinforcement (domestic felony frameworks). In other words, the legal tension at stake is not merely compliance with international human rights law but the hierarchy of values underpinning EU border management. If border integrity consistently supersedes the core values of life and human dignity, the doctrinal architecture of human rights law is functionally subordinated to security rationales.

Accordingly, the Chios incident should be analyzed not only as a maritime tragedy but as a constitutional stress test for the EU human rights regime. The decisive issue is whether accountability mechanisms, domestic courts, the ECtHR, EU oversight bodies, will be able to effectively recalibrate operational practice toward a life-preserving baseline or tacitly normalize deterrence-driven fundamental rights erosion.

Conclusion

Greece exemplifies a broader European paradigm: a so‑called “success story” for deterrence‑based migration control, yet in reality a humanitarian catastrophe for those seeking protection. The country’s migration policies mirror a wider EU strategy that prioritizes border fortification over human life. Hence, a rights‑first recalibration is urgently required. This entails independent and transparent investigations into all reported maritime incidents, such as the full public release of the Chios and Pylos footage, and unhindered support for NGOs engaged in SAR operations, paired with the domestic decriminalization of humanitarian assistance to migrants. In the absence of these measures, the prevailing doctrine of “prevention at all costs” will perpetuate watery graves, turning the Mediterranean into an open cemetery and rendering the protection of migrants’ fundamental rights mere eulogies in default.

Robert Junqueira [Executive Coordinator of the Research and Scientific Careers Bureau of the Research Centre for Justice and Governance (JusGov)] [1]

As AI systems are developed and used by a wide range of individuals and organisations – not least military bodies, as recent events in Iran attest[2] –, it can become unclear who is responsible when something goes wrong. At its core, the debate surrounding responsibility for harm caused by a system (biological or otherwise) with a fractured or nonexistent legal personality is not unprecedented. Well before the age of algorithmic governance, legal and moral reasoning laid considerable groundwork for determining liability under circumstances wherein the link between intent and outcome is obscured by technical artefacts, chains of command, organisational setups, and status-based asymmetries.

In ancient Rome, for instance, legal issues around agency and liability were frequently addressed, prompting the legal order to evolve and respond with gradually emerging solutions. While not necessarily providing us with ready-made schemes, such precedents nonetheless draw our attention to the fact that legal issues involving responsibility have traditionally arisen and remedies were found as a result of incremental steps rather than by means of abrupt, one-off changes. This fact, the problems faced by our ancient peers, and the ways in which they managed them, offer valuable lessons and useful models when tackling today’s pressing AI regulatory challenges.[3]

Customs decay and inventions flourish, but the vexations of legal inquiry do but slumber to wake again. Now, they are stirring anew, propelled into everyday life by algorithms. In response to an emerging concern that transcends national borders, the European Union has positioned itself at the forefront of the regulation of so-called Artificial Intelligence. This pioneering stance is reflected in the formulation of a regulatory landmark (the AI Act) which, by proposing a regulatory architecture that could serve as a stepping stone worldwide, seeks to harmonise legislative frameworks within the EU and to leverage transnational cooperation.

The AI Act looks to make navigating the landscape of responsibility of the hypercomplex society of our day somehow manageable,[4] yet it does so within narrow boundaries, excluding systems as long as they are developed, placed on the market, put into service, or used exclusively for military, defence, or national security purposes (Art. 2; Recital 24). This exclusion is dynamically determined: a system originally intended for civilian use may fall outside the scope of the Act as a consequence of being repurposed for wholly military purposes, whereas a system originally designed for military use may fall within the scope of the Act as a result of being repurposed for civilian, law enforcement, humanitarian, or other purposes the Act does not exclude. In cases of mixed purposes, the Act covers the non-excluded one(s).

While what is regulated by the Act is fairly clear and the regulation is pursuant of Art. 4(2) of the Treaty on European Union (TEU), uncertainty about who is responsible when something goes wrong arises due to the way responsibility lies with a broad network of roles (providers, deployers, importers, and distributors), obligations triggered by the functional role, risk profile, and use of systems (rather than intent), and duties that may be front-loaded at the outset or extend across their lifespan (transparency, oversight, monitoring, and reporting). Altogether, factors like these give rise to an intricate, hard-to-parse horizon of responsibility. The AI Act itself acknowledges and structures this knottiness by mapping roles across a regulatory network and by tying duties to objective risk categories and use cases.

The AI Act addresses these issues by defining the regulated parties and their roles throughout the network (Arts. 2-3: providers, deployers, importers, distributors), categorising systems based on objective risk levels and use cases (Art. 6 and Annex III; prohibitions in Art. 5), stipulating role-specific obligations (Arts. 9-15 as high-risk requirements; Art. 16 as provider duties; Art. 26 for deployers; transparency duties in Arts. 13 and 50; and GPAI provider duties in Art. 53) and building governance and enforcement instruments to operationalise responsibility (conformity assessment and notified bodies in Arts. 30-43; EU database for high-risk systems in Art. 71; national authorities and the AI Office in Arts. 64-70; penalties in Art. 99).

Even with this framework in place, uncertainties can remain regarding roles, triggers, and accountability. While the AI Act successfully provides greater legal clarity on roles and duties, the practical allocation of responsibilities amid evolving regulatory and technological landscapes, multiple parties and their overlapping obligations, and the attachment of duties to objective categories rather than intent—i.e., duties that attach because a system falls within a regulated type or use case, not because a person meant to cause harm—can still produce significant levels of unpredictability, especially as compliance is assessed through a highly complex, deeply technical, and quickly morphing evidentiary record involving possibly very large numbers of parties and the various stages of the lifecycle of the relevant system(s).

Regulation of AI is sorely needed. Good rules are far more than an exercise of power; they are also a means of holding power to account and fostering trusting relationships. In today’s AI ecosystem, clear expectations about responsibilities are needed by the many different actors involved, ranging from developers and companies that deploy systems to people who use them or are otherwise affected by their self-directed functionality. And, because such sophisticated technologies frequently defy geographical and institutional boundaries, a shared baseline for AI governance is all the more critical.

The AI Act is an attempt to preserve relational trust and public accountability while enabling innovation under a framework that remains faithful to the constitutional values of the EU. Framed as a regulation stipulating standardised guidelines for AI and championing human-centric and reliable systems, the Act conveys the message that the bloc’s strategy is not only about imposing restrictions, but also about fostering prudent growth. This development is, moreover, consonant with the spirit and desideratum of the Union: to exert global influence through soft law, that is, to act as a normative power. Such an aspiration does not entail the assertion of universal validity, but rather the provision of a foundation and an incentive for the development of strategies and solutions tailored to the specific circumstances and needs of diverse configurations within the global socio-political and economic fabric.

Inevitably, this presupposes, as it were, an EU-specific repertoire of political values, comprising a discourse on democracy, the rule of law, and fundamental rights (as enshrined in Art. 2 TEU). These values are not abstract or purely legalistic notions, but are grounded in moral reasoning and thoughtfully designed to promote overarching societal prosperity. And yet this is not, nor does it need to be, a claim that the EU offers the only or most legitimate or desirable way of life, or that other civilisational and ideological narratives are misguided or substandard. Standing on a more humble footing, the EU simply has to acknowledge that it is well placed to provide an auspicious reference point in times of global uncertainty.

Reaching as wide a consensus as possible globally would indeed be advisable, but the fact remains that, regardless of how other powers handle such an opportunity, the EU is duty-bound to properly govern the flow and impact of border-crossing technologies within EU territory. In doing so, it must remain true to its own values. This means, at the very least, setting out clear expectations about what counts as acceptable design and deployment, what kinds of uses are off-limits, and what safeguards must be in place when systems operate in ways that can materially affect people’s lives. It also means making those expectations legible across the whole ecosystem, from developers and providers to deployers, public authorities, and end users, so that the same baseline can guide practice, oversight, and contestation.

Questions of responsibility invariably arise in this context. A community that prioritises values such as dignity, fairness, inclusivity, and due process also needs to define a strategy and take action to address any threats or damage to their integrity. In other words, precepts must perforce be transmuted into our shared grammar of obligations, constraints, and redress, instead of just floating above conduct like clinquant finery. If we want those values to matter in real cases, we must be able to unambiguously spell out what concrete safeguards were owed, who owed them, and to whom they were owed. Consider, for example, credit scoring and profiling, i.e., algorithmic assessments that predict creditworthiness or categorise people according to risk. These examples make the point bluntly: without a clear duty to disclose information and a duty-holder to be held accountable, the language of fairness and due process can but be lipstick on a pig.[5]

We need an adequate grammar for attributing responsibility that builds on our baseline of values and takes into account conduct, position, and the risks inherent in certain undertakings. Bearing this in mind, the approach taken by the EU seems less improvised or a mere reaction to isolated problems and closer to a response to a structural governance gap. Indeed, the EU’s regulation of AI signals an official response to the growing fragility of traditional systems of behavioural governance and to the unease that emerges when the settled and time-honoured legal compass of principles is confronted with the novel questions raised by the accelerated and pervasive process of digitalisation through which we are currently passing. This erosion is indicative of a disturbance to a status quo oriented towards efficiency, wherein prudence was assumed to be intrinsic to the human factor.

As so-called intelligent technologies increasingly distinguish themselves through self-directed behaviour, triggering complex execution processes without synchronous supervision by a conscious personality and producing effects that may be profound and detrimental at the level of fundamental rights, it becomes apparent that we are called upon to promote a thorough moral scaffolding and to equip ourselves with a robust and up-to-date legal order. With a view to exercising critical control over the emergence of AI, it falls to moral science to safeguard the justification for the intangibility of the exclusively human nature of responsibility; it is then incumbent upon legal science to draw upon this foundation and to preserve the stability of a minimum common ethical core.

Legal studies in this field must always remain attentive to the productive dialectic between these two spheres. This entails an ever-deeper understanding of the discrepancy between technical compliance and moral capacity, encompassing both what the legal order may seek to enforce and what belongs to the inalienable realm of responsibility. Legal research in this context is thus indispensable and ought to remain aware that computational reasoning is devoid of moral intent, insofar as it lacks free and autonomous will, leaving no room for transgression or transcendence: a machine cannot falter when operating due to faith, love, honour, or mere fancy. Preserving the sense of difference between the human and the machine will be essential if we are to perpetuate a culture of morally grounded discernment, shielding it from being crushed by the autotelic and heavy machinery of a technocentric and, by extension, dehumanised civilisation.

An urgent need to invest in EU Law AI-centred research is most evident in the judicial context. The AI Act already takes this approach by imposing transparency, record-keeping, and human oversight duties. Nevertheless, translating these duties successfully into judicial practice calls for sustained doctrinal and methodological efforts. In view of the amorality of AI, it is above all imperative that the EU Law scientific community address every scenario in which automated data processing may arise as the cornerstone of the judicial function, continuously building on the responses we have been devising to address the question of whether it is desirable for the final verdict in matters of justice to remain a product of human deliberation.

For the sake of caution, it is critical to recognise that judicial decision-making cannot be reduced to patterns—that is, the average quantum—of past behaviours collected and systematised from datasets algorithmically generated and processed. In adjudicating a real-life case, it is a defining prerogative of the judiciary to attempt to foresee the future of the persons to whom its decision is addressed. This is a personal and indivisible future, which need not correspond to the statistically median outcome derived from prior anticipations of the lives of countless other individuals, with their undeniable specificities, and which will inevitably differ from the singular recipients of the jurisdictional provision in the case at hand.

On this and other related issues, the path signposted by the EU is one of openness to moderation in the face of the current feverish enthusiasm for technological progress. Nonetheless, we should not halloo before we are out of the wood. Only if properly explored will this openness make it possible to reap the benefits of technological innovation without compromising the human countenance and the foundational premises that sustain us as a community governed by law. It is now up to R&D Units such as the Research Centre for Justice and Governance of the School of Law of the University of Minho, steadfast in their commitment to engaging meaningfully with peers in their own and other fields of study, to undertake the labour-intensive task of, besides developing principled responses to the national-exclusive regulatory space left outside the AI Act, leveraging the Union’s openness and preventing it from turning into a principium sine fructu.

[1] More information and contact: https://www.robertjunqueira.com/. This set of reflections arose from dialogues with both the Coordinator of the Research Group for Studies on European Union Law of the Research Centre for Justice and Governance of the School of Law of the University of Minho, Professor Pedro Madeira Froufe, the President of the Ethics Committee of the Polytechnic University of Cávado and Ave, Professor Irene Portela, and my colleague Beatriz Melo. Further to this, this is a result of a kind invitation extended by an eminent scholar of European Union legal studies, Professor Alessandra Silveira.

[2] See Namir Shabibi and Alex Croft, “AI, a dead student, and US airstrikes: how a civilian became caught up in a new age of warfare”, The Independent, March 10, 2026, https://www.independent.co.uk/news/world/middle-east/ai-airstrike-civilian-killed-us-centcom-iraq-anthropic-b2926712.html.

[3] See, e.g., Klaus Heine and Alberto Quintavalla, “Bridging the accountability gap of artificial intelligence: what can be learned from Roman law?” Legal Studies 44 (2024): 65–80, https://doi.org/10.1017/lst.2022.51.

[4] To find out more about the hypercomplex society and why it may be impossible to govern, read Piero Dominici, Beyond black swans: inhabiting indeterminacy (Cham: Springer, 2026), https://doi.org/10.1007/978-3-032-09029-4.

[5] See Alessandra Silveira, “Automated individual decision-making and profiling [on case C-634/21 – SCHUFA (Scoring)],” UNIO – EU Law Journal 8, no. 2 (2023): 74–85, https://doi.org/10.21814/unio.8.2.4842.

Picture credit: by Tara Winstead on pexels.com.

The post Italy’s referendum – is the Italian right as united as it seems? first appeared on LSE European Politics.

Tamta Gventsadze, PhD candidate in law, UNITUS

Photo credit: Mstyslav Chernov, via Wikimedia Commons

Introduction

On February 5, 2026, the Court of Justice delivered its judgment in Case C-718/24, Aleb concerning the interpretation of Articles 33(2)(c), 38 and 46 of Directive 2013/32 in the context of the ‘safe third country’ concept and the right to an effective remedy, Article 47 of the Charter of Fundamental Rights of the European Union.

This analysis examines the Aleb judgment in light of the ‘safe third country’ concept under Directive 2013/32 (Asylum Procedures Directive, APD) and its replacement by Regulation (EU) 2024/1348 (Asylum Procedures Regulation, APR) from 12 June this year. After mapping the factual background and the Court’s clarification of the cumulative safeguards governing the presumption of safety, it then considers the impact of the 2026 amendments and assesses their implications for judicial control and fundamental rights protection.

I. Facts of the Case

The applicant, NP, is a Syrian national and an unaccompanied minor who lodged an application for international protection in Bulgaria on 2 November 2023. During the interview conducted on 1 December 2023, he stated that he had lived in Aleppo (Syria) and had left two to three months earlier with his brothers because of the war. Before “illegally entering the Bulgarian territory” (para. 19), NP stayed in Türkiye for approximately one month, where his brothers remained with three of his sisters who already lived there.

By decision of 18 June 2024, the Chairperson of the Bulgarian National Refugee Agency rejected the application, refusing to grant him both refugee and humanitarian status. The authority accepted that Syria was affected by internal armed conflict and indiscriminate violence and acknowledged that the applicant was exposed to a real threat to his life or person. Nevertheless, it refused to grant protection on the ground that Türkiye constituted a ‘safe third country’ in which the applicant could seek protection. The decision relied, inter alia, on the fact that the applicant had already lived in Türkiye for about a month without suffering harm, had close family members there, and that Syrian nationals in Türkiye benefited from temporary protection and protection against forced return; finally, “their basic needs [were] satisfied” (para. 20).

The referring court expressed doubts as to the compatibility of this approach with Directive 2013/32, especially considering the absence of both a defined methodology for applying the ‘safe third country’ concept and the lack of explicit procedural guarantees under Bulgarian law allowing the applicant to challenge the existence of a sufficient connection with Türkiye (paras 24-26).

II. The Presumption of Safety under Article 38 of the Directive

The judgment deals with the legal nature and limits of the presumption underlying the ‘safe third country’ concept. The Court begins by recalling that the application of Article 33(2)(c) of the APD (ie, providing that asylum applications are inadmissible where the ‘safe third country’ principle is applied) is conditional upon strict compliance with the requirements of Article 38 thereof (ie the definition of ‘safe third country’ and the conditions related to it). In para. 48 of the judgment, it is expressly stated that the conditions laid down in Article 38 are cumulative, with the result that the inadmissibility ground cannot be applied where any one of those conditions is not satisfied.

This formulation makes clear that the presumption of safety is neither automatic nor self-standing, it is legally constructed and constrained by a series of substantive and procedural safeguards (para. 46). More precisely, Article 38(2) requires Member States to regulate the safe third country concept through national law and that the national rules ensure: (i) there is a sufficient connection between the applicant and the third country so that return there is reasonable; (ii) must define a methodology for applying the concept, which includes either a case-by-case safety assessment or the designation of generally safe countries; (iii) must guarantee an individual examination and allow the applicant to challenge both the safety of the third country in their specific circumstances and the existence of the required connection.

Therefore, the key requirement in the safe third country designation is the existence of a “connection” between the applicant and the third country. The Court emphasizes that Article 38(2)(a) obliges Member States to define in national law criteria enabling authorities to determine whether such a connection exists and whether return to that country is reasonable (paras 51–52). Since the Directive does not define “connection,” Member States retain discretion to specify the criteria, but within EU limits.

Importantly, the Court further reiterates its prior case law that mere transit through a third country cannot, on its own, justify the conclusion that return there is reasonable (para. 54). This statement substantially narrows the presumption. It prevents Member States from relying on minimal factual links and requires a qualitative assessment of the relationship between the applicant and the third country, considering factors such as duration and circumstances of stay and family ties. It is evident that the presumption cannot be based merely on the “transit” criterion. Even where national law relies on the notion of “stay,” national courts must assess, in light of all circumstances, whether that stay genuinely establishes a sufficient connection (para. 55).

The Court also addresses national lists of safe third countries. Member States may, in principle, designate safe third countries by general act. However, such designation does not dispense with the obligation to conduct an individual assessment. National law must provide a methodology for a case-by-case evaluation of both the country’s safety for the applicant and the existence of a sufficient connection (para. 65). The presumption must remain rebuttable, and the applicant must be able to challenge the existence of that connection.

The judgment firmly situates the safe third country concept within the framework of effective judicial protection. Article 38(2)(c) must be read in conjunction with the ‘effective remedy’ rights in both Article 46 of the Directive and Article 47 of the Charter (paras 69-74). Even if national law does not explicitly confer such power, a court hearing an appeal must verify whether a sufficient connection exists.

Accordingly, the Court recalls that Article 46(1) of the Directive guarantees a right to an effective remedy in asylum cases and that Article 46(3) requires a full and ex nunc examination of both facts and law in asylum appeals. This standard applies even in inadmissibility cases and does not necessarily require a substantive assessment of protection needs, but it does require full judicial scrutiny of admissibility conditions. Furthermore, Article 47 of the Charter enshrines the principle of effective judicial protection and is directly applicable, thereby requiring national courts to conduct comprehensive review consistent with EU fundamental rights standards.

Therefore, when reviewing a decision declaring an application inadmissible on ‘safe third country’ grounds, national courts must conduct a full and up-to-date examination of whether the third country is safe for the applicant and whether all cumulative conditions, including the connection requirement, are fulfilled (para. 75). The Court thus subjects the presumption of safety to meaningful judicial scrutiny grounded in Article 47 of the Charter.

Finally, the Court confirms that an application may be declared inadmissible on ‘safe third country’ grounds even where the applicant faces a real risk of serious harm in the country of origin (ie, as distinct from the ‘safe third country’). This confirms that the ‘safe third country’ concept functions as a jurisdiction-allocating mechanism rather than as a substantive denial of risk in the country of origin. Precisely because it allows displacement of protection despite such risk, strict compliance with Article 38 safeguards is imperative.

III. Relevant changes of legislative framework

It is further necessary to underline that while the Court has provided meaningful clarifications regarding safe third country concept and judicial protection in light of corresponding provisions under Directive 2013/32 in a few months’ time this instrument will be replaced by another secondary EU legislation in the form of Regulation (EU) 2024/1348 of the European Parliament and of the Council of 14 May 2024 establishing a common procedure for international protection in the Union and repealing Directive 2013/32/EU (already discussed in great detail here).

Upon a comparative assessment of these two contrasting instruments, several illustrative differences emerge between the provisions concerning safety presumptions, and those of appeals.

Article 33 of the APD addresses inadmissible applications, framing them as an exception to the obligation to examine the substance of an application, meaning that Member States are not required to assess the merits of qualification for international protection where one of the exhaustively listed grounds for inadmissibility applies. The list is limited: Art. 33(2) includes protection granted by another Member State, first country of asylum, ‘safe third country’, subsequent applications without any new elements, and applications lodged by dependents who had consented to be included in another application.

Article 38 of the APR restructures this framework, by separating the decision on admissibility from the decision on the merits. Article 39(3) explicitly governs rejection as unfounded following substantive examination pursuant to the ‘Qualifications’ Regulation (EU) 2024/1347. This structural clarification seems to strengthen the distinction between inadmissibility and unfoundedness, which had already been emphasized by the Court.

In addition, relevant to the current analysis, a notable development concerns the first country of asylum and ‘safe third country’ grounds of inadmissibility. In the APD, Article 33(2)(b) and (c) refer to Articles 35 and 38 without expressly conditioning inadmissibility on the likelihood of admission or readmission. The Regulation introduces an explicit safeguard, with Article 38(1)(a) and (b) essentially stating that inadmissibility may apply unless it is clear that the applicant will not be admitted or readmitted to the third country. Therefore, changes regarding admissibility grounds combine expansion with procedural tightening and partial codification of judicial safeguards.

To further detail the elements regarding the concept of ‘safe third country’, it seems clear that the evolution of the rules on this principle from Article 38 of the APD to Article 59 of the 2024 Regulation, especially as amended in 2026, reflects an apparent shift in structure and scope, even if several aspects remain unchanged.

First off, at the level of safety criteria, both instruments require absence of threats to life or liberty on Convention grounds, absence of serious harm, respect for non-refoulement, and protection against removal contrary to international law. The Regulation clarifies the content of “effective protection” through reference to Article 57 and to Regulation (EU) 2024/1347. As defined by Article 57, effective protection can be attained in a country that has ratified and respects the Geneva Convention on Refugee Status, within any permitted reservations or limitations. However, where a geographical limitation applies, or where the Convention has not been ratified, protection must be assessed against minimum criteria of permission to remain on the territory, access to sufficient means of subsistence to ensure an adequate standard of living, access to healthcare and essential treatment, access to education under general national conditions, and the availability of protection until a durable solution is found. Interestingly, these minimum criteria resemble those of subsidiary protection guarantees, but the context refers to international protection.

Secondly, additional significant changes concern Union and national designation mechanisms introduced by the Regulation. Article 59(2) and (3) of the Regulation, which have no previous equivalent in the Directive, allow partial designation for specific territorial parts or identifiable categories of persons and require reliance on a broad range of sources. Importantly, Art. 59(5)(b) explicitly maintained, that the ‘safe third country’ concept may only be applied where “there is a connection between the applicant and the third country in question on the basis of which it would be reasonable for him or her to go to that country” (the same standard of ‘reasonable’ connection as employed by the Court in Aleb).

However, the 2026 amendments substantially reshape the connection criterion, by removing it as a mandatory condition and introducing two additional bases. The concept may now apply where the applicant transited through the third country (explicitly opposing to what the Court prohibits in Aleb). It may also apply where an agreement or arrangement exists requiring the third country to examine protection claims. As already wittily named elsewhere, this ‘Rwanda clause’ permits transfer without prior connection or transit. Moreover, the removal of automatic suspensive effect in safe third country appeals in the 2026 amendments seem to further intensify a restrictive shift, although there is still a possibility for requesting a judicial suspension.

The most problematic aspect seems this possibility of transferring an individual to a state where they might have never even been to, but the secondary law would permit it solely due to the existence of an agreement (often in non-legally binding form of MoUs) between a EU Member State or the Union and possibly any third country. It is true that an applicant will maintain a possibility to appeal this decision, however, they might have to first ask for a suspensive effective of the appeal, which further complicates an already arduous procedure. In theory, a person might end up being transferred to such a supposedly ‘safe’ third country before a decision is made upon their request to remain, which could create a rather unclear legal consequence; would a person potentially have to be brought back, shall their appeal succeed? It is true that this particular scenario might be extremely rare in practice, but does this rarity allow for its legality?

Nevertheless, despite this expansion in scope, the APR maintains several core safeguards, explicitly stating that admission or readmission must be ensured and that individual assessment remains required (which could become more restricted in practice). Special guarantees also apply to unaccompanied minors, including best interests and prior assurances of protection; and the ‘Rwanda’ clause cannot apply to unaccompanied minors at all.

As confirmed by the Alace judgment, designations of ‘safe countries of origin’ must remain subject to judicial review under Article 47 of the Charter. The Court held that national courts must be able to examine compliance with material designation criteria and to rely on independent sources of information. This reasoning applies by analogy to ‘safe third country’ designations and considering that the Charter has the same legal value as primary EU law, legislative attempts to narrow judicial review cannot override it.

In this context, the Aleb judgment constitutes an additional message. It reaffirms that application of the ‘safe third country’ concept is subject to cumulative conditions and full judicial scrutiny. Even as the legislature broadens the concept and limits suspensive effect, the Court insists on effective judicial protection and strict assessment of safety criteria.

IV. Concluding Assessment

The Aleb judgment does not abolish the presumption of safety inherent in Article 38 of the Directive. However, it subjects the presumption to cumulative substantive conditions, mandatory individualized assessment, and full judicial review.

In more detail, the ECJ allows that the determining authority may apply the concept of ‘safe third country’ on the basis of information from publicly available sources and rely on a national list of safe third countries, if such exists, but this is only provided that national law also defines the methodology applicable for assessing, on a case-by-case basis, according to the particular circumstances of the applicant for international protection, whether the third country concerned satisfies the conditions for being regarded as safe for that applicant and the possibility for that applicant to challenge the existence of a connection, within the meaning of Article 38(2)(a).

Moreover, the Court’s insistence, particularly in paragraphs 48, 54 and 65 of the judgment, on the cumulative nature of the safeguards, the insufficiency of mere transit, and the necessity of a defined methodology, significantly limits the discretionary space of national authorities. The ECJ reinforces the procedural containment of the ‘safe third country’ mechanism, by linking these cumulative requirements to Article 46 of the Directive and Article 47 of the Charter, and by requiring national courts to verify the existence of a connection even where national law is silent in this regard.

Therefore, the Court ties the existence of the ‘safe third country’ presumption to two co-existing elements: clearly defined methodology underlining individual assessment (which includes the existence of “reasonable” connection between the applicant and the safe third country) and a possibility for judicial review of the connection requirement. It seems evident from this judgment, that the safety presumption, be it national or supranational level, would otherwise be invalid. In a way, Aleb strengthens the doctrinal link between inadmissibility decisions and effective judicial protection. The presumption of safety is permitted, per se, but only as a structured, reviewable and rebuttable legal construction embedded within the broader guarantees of EU fundamental rights law.

The combined effect of Alace and Aleb indicates that the Court of Justice does not seem to be prepared for relaxing the standards governing the designation and review of ‘safe third countries’. Nevertheless, it remains to be seen whether and how the Court will respond to the legislative changes, considering that they aim to abolish the mandatory connection element and restrict safeguards when challenging it.

The post Europe’s Ukraine consensus still holds – but the far right is splitting over Russia first appeared on LSE European Politics.

The post What skills policy can tell us about the tunnel vision of policymakers first appeared on LSE European Politics.

Alessandra Silveira [Editor of this blog, Coordinator of Jean Monnet Centre of Excellence “Digital Citizenship and Technological Sustainability” (CitDig), University of Minho]

In the week marking the fourth anniversary of the invasion of Ukraine on February 24, 2022, a new book by Francisco Pereira Coutinho, Professor of public international law and European constitutional law at NOVA University Lisbon, was launched in Portugal. I had the pleasure of presenting the work on February 26, together with Pedro Froufe, editor of this blog. Right at the beginning of the book, entitled “Guerra, mentiras e direito internacional”[1] – which directly translates to English as “War, lies and international law” –, the author explains what his main motivation was: he wanted to tell the story of the invasion of Ukraine from the perspective of an expert in international law. He then gathered the questions that his colleagues and journalists had asked him – as he is also a commentator for CNN Portugal – and the result is, in my view, a compelling and courageous manifesto in defence of international law.

As the metaphor on the book cover says, international law does not cease to exist because it is violated, just as grammar does not disappear because someone writes poorly. However, continuous violations of the rules can render them meaningless. This is because law (in general) is an abstraction – it only works when those subject to the rules recognise their legitimacy and comply with them, above all because they understand that it is worth maintaining peace, security, order and justice. But in the (specific) field of international law, the weaknesses of the law are more evident, because the recipients of the rules are sovereign states, which, if they have sufficient strength to do so, simply abandon their moral compass and cease to comply with the agreed rules – and who is going to rise up against brute force?

In this scenario, international law loses any effective normative force, being subverted as an instrument serving geopolitical purposes. Moreover, international law has been systematically invoked by Russia to (supposedly) justify and legitimise its acts of violence. This is precisely why Francisco Pereira Coutinho’s book is so important, to ensure that the “law of the jungle” does not prevail in international relations, not even (and especially not) in times of war.

As (then) Advocate General Poiares Maduro recalled in his conclusions in the Kadi case before the CJEU,[2] “it is when the cannons roar that we especially need the laws”. Moreover, it is the rules of international law that have legitimised military aid to Ukraine, aimed at repelling Russia’s brutal aggression – which has now been going on for four years, against all expectations. That is why Francisco Pereira Coutinho does not mince his words: he argues that the war in Ukraine is a war of conquest and annexation, in flagrant violation of the United Nations Charter. There is an aggressor (Russia) and there is a victim of aggression (Ukraine) – and this must be stated emphatically to those who seek to reward the aggressor by recognising territorial annexations by force.

But beyond international law and war, the title of the work has a third theme: lies. This is because lies have always been a weapon of war. So much so that Hannah Arendt, in her text on “truth and politics”,[3] argues in defence of factual truth, drawing on the question of culpability for the outbreak of the First World War. Although the blame for World War I was a controversial topic in 1920, no one would have said that Belgium invaded Germany. To erase from collective memory the fact that on the night of 4 August 1914, German troops crossed the Belgian border, it would have been necessary to have a monopoly of power over the entire civilised world. However, Hannah Arendt warned that such a monopoly of power is far from inconceivable – and Arendt could not even have predicted the disinformation on social media…

Of course, opinions may differ widely; it is equally true that each generation can reinterpret the facts – or rearrange the facts according to its own perspective. But no one has the right to subvert the factual truth – and the factual truth is that there was an invasion of Ukraine. That is why Francisco Pereira Coutinho separates “facts” from “propaganda” in the war in Ukraine, to identify fabrications and manipulations. The author confronts allegations with verifiable facts and applicable international law. And what methodology was adopted for this purpose?

The method was inspired by Alan Dershowitz’s work on the Israeli Palestinian conflict[4] and is basically based on what follows. Each chapter opens with a question, followed by an allegation made by one of the belligerents – almost always Russia. This allegation is supported by statements from relevant actors in the context of the war. Finally, the allegation is refuted by the reality of the situation, on the basis of sound legal reasoning, in accordance with applicable international law. The result of this endeavour is appealing, as the discourse is understandable even to those unfamiliar with law, were it not for the author’s background in teaching, which enables him to simplify what is complex. Therefore, the narrative that the author presents to us, namely the story of the invasion of Ukraine, is woven together through questions and answers.

Forty-six questions were selected, and I will take the liberty of identifying some of my favourites: Are Russians and Ukrainians one and the same people? Is Crimea Russian? Can Russia’s annexation of Crimea be recognised? Why did the Minsk Agreements fail? Is NATO enlargement an existential threat to Russia? Could the separatist republics of Donbas secede from Ukraine? Could Russia intervene in self-defence? Was there a duty to intervene on humanitarian grounds? Do the self-determination referendums legitimise the Russian annexations? Can Ukraine attack and occupy Russian territory? Can Ukraine be supported militarily? Can the European Union ban Russian media? Can frozen Russian assets be transferred to Ukraine? Can Russia be expelled from the Security Council? Has Russia been expelled from the Council of Europe? Can the International Criminal Court try Russian citizens? Can Putin be arrested under the International Criminal Court warrant? Can Putin be tried in a special court? Is Russia committing genocide? Is the Russian minority being discriminated against?

In any case, to pique the interest of Portuguese-speaking readers – and show some generosity to those who cannot read Portuguese – I will focus on three issues that I consider essential:

1) The first, discussed in chapter 34 of the book, asks the following question: “Was there a massacre in Bucha?”.[5] This chapter clearly reveals the extent to which, during war, violence carries with it an additional element of arbitrariness – as Hannah Arendt had already said. Moreover, in no other scenario does good or bad luck (Fortune, as Machiavelli put it) play such a decisive role in human affairs.[6] To answer the question, “Was there a massacre in Bucha?”, the author begins with the following allegation: “The massacre in Bucha was nothing more than a propaganda stunt carried out by Ukraine with the aim of derailing the peace negotiations with Russia that were taking place in Istanbul.” This claim is followed by a number of statements, among which I have selected the one made by the Ministry of Defence of the Russian Federation on 3 April 2022: “The photos and videos from Bucha are yet another episode staged by the Kiev regime for the Western media, just as happened in Mariupol with the maternity hospital, as well as in other cities.” However, the reality is as follows: “The Russian army tortured and summarily executed dozens of people in Bucha in March 2022.” The reasoning is lengthy, but I will reproduce a few excerpts: “The Russian army reached the gates of Kyiv. Between 5 and 30 March 2022, it occupied Bucha, a suburb located five kilometres from the Ukrainian capital. Around 5,000 civilians remained in the besieged city. Upon entering liberated Bucha on 2 April 2022, the international press reported the existence of dozens of decomposing corpses in the open air. Many of the bodies showed signs of shooting and torture. In the following months, Ukrainian authorities recovered a total of 422 bodies, most of them exhumed from mass graves and makeshift graves. The Independent International Commission of Inquiry on Ukraine has documented in detail dozens of summary executions in Bucha. (…) The massacre in Bucha was not an isolated case. It is part of a pattern of systematic violence that characterises Russian military conduct, marked by summary executions, torture, forced labour, arbitrary detentions, forced displacement and indiscriminate and disproportionate attacks on civilians and civilian property, which caused, between 24 February 2022 and the beginning of June 2025, at least 13,580 civilian deaths – including 716 children – and 34,115 civilian injuries, including 2,173 children. (…) The European Court of Human Rights, which examined only violations that occurred up to 16 September 2022, highlighted in particular the Russian attacks on maternity hospital no. 3 and on hundreds of civilians sheltering in a clearly marked theatre in Mariupol in March 2022, as well as on a crowd of civilians being evacuated at the Kramatorsk railway station in April 2022. In 2024, the International Criminal Court issued arrest warrants for four senior Russian military officers (…) for war crimes and crimes against humanity, for allegedly ordering deliberate attacks on civilians and civilian property in Ukraine, including critical energy infrastructure.”

2) The second issue that I consider essential is the one addressed in chapter 42, which raises the following question: “Can Ukraine join NATO and the European Union?”. This question is particularly relevant to readers of this blog, because it presents a future perspective for Ukraine within the context of an integrated Europe. To answer the question, “Can Ukraine join NATO and the European Union?”, the author begins with the following allegation: “Ukraine cannot join NATO or the European Union while it is involved in a territorial conflict with Russia.” This assertion is followed by a number of statements, from which I have selected the following by Viktor Orbán on 25 June 2025: “If the European Union accepts Ukraine as a member, this will result in open warfare at this point in time and, after a ceasefire, will entail a continuing risk of war between Europe and Russia.” However, the reality is as follows: “The Russian occupation does not in itself constitute an obstacle to Ukraine’s accession to NATO and the European Union. Germany joined NATO despite part of its territory being occupied by the Soviet Union. Cyprus is part of the European Union despite part of its territory being occupied by Turkey.” The reasoning is once again lengthy, so I will reproduce only an excerpt: “The fact that Ukraine has a territorial dispute with Russia, which is illegally occupying several Ukrainian regions, does not in itself constitute a legal obstacle to Ukraine’s accession to NATO or the European Union. The internationalisation of the conflict can be avoided if the mutual defence clauses provided for in Article 5 of the Washington Treaty and Article 42(7) of the Treaty on European Union are expressly limited to the territory under the effective control of the Ukrainian state, similar to the solutions adopted, in different contexts for Germany and Cyprus. This possibility, however, requires the stabilisation of the battlefront, which can only be achieved – as was the case in Cyprus and Germany – through a ceasefire (de facto or de jure) or, ideally, through the conclusion of an armistice agreement.”

3) Finally, the third issue that I consider crucial is the one raised in chapter 46, which asks the following question: “Will Russia trade peace for territory?”. This chapter is noteworthy because the author envisages peace, but explains to what extent the recognition of the “new territorial realities” – that is, Russia’s annexation of the Ukrainian regions of Crimea, Donetsk, Luhansk, Kherson and Zaporizhia – would not be sufficient for a lasting cessation of hostilities, as for Putin this does not eliminate the “root causes of the conflict”, condensed into the triad of “neutralisation, demilitarisation and denazification” of Ukraine. To answer the question, “Will Russia trade peace for territory?”, the author begins with the following allegation: “There will be peace when Ukraine recognises Russian sovereignty over the «new territorial realities».” This allegation is followed by several statements, among which I have selected one made by Vladimir Putin on 14 June 2024: “As soon as they declare in Kiev that they are not ready to (recognise the new territorial realities) and begin an effective withdrawal of troops (from the occupied regions), and also officially declare that they are abandoning their intention to join NATO, we will immediately give the order to cease fire and start negotiations – immediately, without delay.” However, the reality is as follows: “Exchanging peace for territory rewards the aggressor and does not resolve the «root causes of the conflict», which are identity-based. It merely postpones the solution to the problem.” The reasoning for this position is also extensive, of which I will share an excerpt: “On the other side of the Atlantic, the Trump administration is considering recognising Crimea in exchange for persuading Putin to agree to a ceasefire in Ukraine. This strategy of exchanging peace for territory will not work, as it only satisfies Moscow’s secondary territorial interest, and not its strategic identity goal of keeping Ukraine within its sphere of influence. (…) The Kremlin demands, above all, that the «root causes of the conflict» be resolved. And what is the main cause of the conflict? The decision of the Ukrainian people to associate themselves with the European Union and to establish themselves as a democratic state governed by the rule of law.”

In the four years since the invasion of Ukraine, it is important to remember that without values, everything is equally valid – as Albert Camus put it. If we believe in nothing, if nothing has meaning, if we cannot affirm any values, then everything is permitted and nothing matters. Therefore, there is neither good nor evil – and Hitler was neither wrong nor right. If we think that nothing has meaning, we must conclude that whoever wins is right. Because if nothing is either true or false, if nothing is either good or bad, then brute force prevails, where there is only room for masters and slaves. To this extent, the freedom we are challenged to achieve these days is that of not lying – for only then will we be able to identify the reasons for living and dying.[7] 

[1] Francisco Pereira Coutinho, Guerra, mentiras e direito internacional (Lisboa: Zigurate, 2026).

[2] Opinion of Mr Advocate General Poiares Maduro delivered on 16 January 2008, Yassin Abdullah Kadi and Al Barakaat International Foundation v Council of the European Union and Commission of the European Communities, joined cases C-402/05 P and C-415/05 P, ECLI:EU:C:2008:11.

[3] Hannah Arendt, Entre o passado e o futuro, 9^th edition (São Paulo: Perspectiva, 2022).

[4] Alan Dershowitz, The case for Israel (Wiley, 2003).

[5] All quoted excerpts of this book have been freely translated by the Author of this post.

[6] Hannah Arendt, Sobre a violência (Lisboa: Relógio d’água, 2014).

[7] Albert Camus, Conferências e discursos (1937-1958) (Porto: Livros do Brasil, 2022).

Picture credit: by Bruno Reynaud de Sousa.

The post Berkay Ozcan: “Immigration is not a silver bullet for falling birth rates” first appeared on LSE European Politics.

The post How the Czech Senate protects the country from democratic backsliding first appeared on LSE European Politics.

Karin de Vries (professor of fundamental rights law, Utrecht University) and Sarah Ganty (JSD candidate, Yale Law School; FNRS Post doc fellow, UCLouvain)

Photo credit: Kristoffer Trolle, via Wikimedia Commons

The so-called Danish ‘ghetto law’ case has been one of the most closely watched judgments of 2025 before the Court of Justice of the European Union (CJEU), long awaited by lawyers working in EU anti-discrimination law and beyond. Its resonance has extended well beyond legal circles, attracting sustained attention from the media and NGOs.

Much has already been written on the judgment, from different perspectives (for example, here, here, here and here). The focus of this contribution is on how the judgment shapes the definition of racial discrimination in EU law, including the emerging tension between objective and social understandings of ethnic origin, the Court’s application of the concepts of direct and indirect discrimination and its reticence to address the structural racism at the roots of the Danish housing policy. Before proceeding to our analysis we first offer a brief outline of the case and the Court’s reasoning.

Background: the Danish ‘ghetto law’

At the heart of the case lies a practice of state-driven gentrification as part of the Danish housing policy. The Danish Law on public housing seeks to reduce the proportion of public housing in certain designated neighbourhoods, officially labelled “transformation areas”. Under the scheme, authorities may sell buildings to private developers, demolish existing housing, or convert family dwellings into accommodation for young people. These measures may entail the unilateral termination of tenants’ leases, resulting in their forced displacement.

To determine which neighbourhoods are subject to restructuring, an earlier version of the Law on public housing introduced the labels of “ghettos” and “hard ghettos”. These labels have since been replaced by the less openly yet still stigmatising terms “parallel societies and “transformation areas”, without however incurring any substantive changes to the scheme. A neighbourhood classifies as a parallel society when it satisfies at least two out of four socio-economic criteria—relating to unemployment, education levels, criminal convictions, and average gross income—and at least 50 per cent of its residents are “immigrants and their descendants from non-Western countries”. Where such a classification persists for five consecutive years, the area is classified as a “transformation area” and becomes subject to the abovementioned far-reaching restructuring measures. The Law on Public Housing also identifies “vulnerable areas”, neighbourhoods that meet the same socio-economic criteria as parallel societies but do not have a majority of “non-Western” residents. These areas are not eligible for restructuring. Hence, the possibility of restructuring turns explicitly on the “Western”/“non-Western” distinction, making the risk of forced relocation contingent upon the demographic composition of a neighbourhood.

The applicants are residents of designated “transformation areas” in the municipalities of Slagelse and Copenhagen. The applicants from Slagelse challenged the termination of their leases, whereas the applicants from Copenhagen sought invalidation of the ministerial decision approving the development plan for their neighbourhood. Both cases reached the High Court of Eastern Denmark, which referred two preliminary questions to the Court of Justice First, whether the criterion “non-Western immigrants and their descendants” constitutes a distinction based on ethnic origin within the meaning of the Race Equality Directive (RED); and second, whether the Danish legislation gives rise to direct or indirect discrimination.

The Judgment

The Opinion of the Advocate General (Ćapeta) answered both questions in the affirmative, concluding that the distinction is based on ethnic origin and that the difference in treatment amounts to direct discrimination, although it could also be qualified as indirect discrimination. While the Grand Chamber does not fully adopt the AG’s Opinion, much of its reasoning is reflected in the judgment. The CJEU, first of all, concurs with the AG that a public housing scheme such as that existing in Denmark is capable of falling within the material scope of the RED (§ 66). It then proceeds to examine whether the Danish legislation constitutes direct ethnic discrimination. While leaving it to the referring court to ultimately decide this issue, the Court offers detailed guidelines which allow very little room for a negative answer. Still, in case the referring court should conclude that there has been no direct discrimination, the CJEU also provides guidelines to determine whether there has been indirect discrimination. Here the Court interprets the requirement of Article 2(2)(b) RED that persons of a racial or ethnic origin must be put at a ‘particular disadvantage’, as well as the criteria for objective justification. The Court leaves to the referring court to decide, notably, if the Danish legislation puts persons belonging to certain ethnic groups at a particular disadvantage (§ 143) and, if that is the case, if the restructuring measures are appropriate, necessary and proportionate in relation to the public interest of promoting social cohesion and integration (§§ 164-165).

On ethnic origin: a step forward but not quite there yet

A central issue in the judgment is whether the category of “non-Western immigrants and their descendants” falls within the prohibited ground of racial or ethnic origin under the RED. The Court of Justice has long displayed an uneasy relationship with the concepts of “race” and ethnic origin — most notably with the former, which it has traditionally been reluctant to invoke expressly. In CHEZ, the Grand Chamber held that ethnic origin ‘has its origin in the idea of societal groups marked in particular by common nationality, religious faith, language, cultural and traditional origins and backgrounds’ (§ 46). While this definition in itself is not unnecessarily restrictive, its application in subsequent cases demonstrated a very restrained understanding of the concept. In Jyske Finans as well as Land Oberösterreich v KV the Court’s focus was on establishing the presence of objective characteristics rather than identifying social processes of racialisation (here) and othering through which such characteristics gain social meaning and become markers of racialized social hierarchy. Moreover, it was suggested that the requirement of a ‘particular’ disadvantage in Article 2(2)(b) RED meant that only ethnically homogeneous groups could seek protection under the Directive, thus excluding differential treatment of heterogeneously composed groups such as ‘third-country nationals’. This approach has been widely criticised in the literature (for example here, here,and here). Commentators have argued that the Court’s understanding of ethnic origin risked hollowing out the protective scope of EU anti-discrimination law and failed to account for the lived realities of racialised exclusion suffered by migrants in particular.

Despite AG Ćapeta’s invitation (Opinion, § 69) the Court does not expressly recognize ‘ethnic origin’ as a socially constructed category. It confirms, however, that the term ‘racial or ethnic origin’ in the RED must be understood in light of the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD) and of Article 14 European Convention of Human Rights (ECHR) and the case law of the European Court of Human Rights (ECtHR). This confirms the Court’s growing tendency to engage with external sources, especially its Strasbourg counterpart, even as EU accession to the European Convention on Human Rights remains pending. The reference to ECtHR case law is moreover significant as that court has previously espoused a less rigid understanding of ethnicity, notably in the Grand Chamber judgment in Biao v. Denmark where it held that ‘Danes of foreign origin’ were subject to ethnic discrimination. The CJEU also recalls the ECtHR’s consistent qualification of racial discrimination as a ‘particularly invidious form of discrimination which, in view of its perilous consequences, requires […] special vigilance and a vigorous reaction’ (§ 79). It thus makes clear that the issue at stake is one of racial discrimination and that no relevant distinction exists in this regard between the concepts of “race” and ethnic origin. Still, it fails to draw any strong doctrinal consequences from this acknowledgement in the form of a more explicitly social conception of racial discrimination.

A very welcome aspect of the case is the Court’s clarification that ethnic origin—as previously defined in CHEZ—need not be determined by reference to a single or homogeneous ethnic group (§§101-104; §§134-140). Returning to its first judgment on the RED, Feryn (§103), the Court clarifies that the notion of ethnic origin may apply to broadly formulated criteria such as ‘allochtones’, ‘foreigners’ or, indeed, ‘non-Western immigrants’. Still, it continues to insist, as it did in Jyske Finans, that a single characteristic such as nationality or country of birth, can never be indicative of ethnic origin (§ 86). Instead, ethnic origin must always be determined based on a combination of factors. As we argued earlier, this insistence on multiple characteristics denies the reality of racial discrimination in which a single feature (such as someone’s nationality, surname or skin colour, for that matter) may be the reason for differential treatment if that feature functions, in the given context, as a marker of “race” or ethnicity. The Court thus keeps open a significant loophole in EU anti-discrimination law—one that Member States may exploit by designing exclusionary measures around formally non-protected criteria, while shielding them from scrutiny under the prohibition of racial and ethnic discrimination.

In the present case, however, the single characteristic requirement does not stand in the way of a finding of ethnic discrimination. The Court is satisfied that the criterion of ‘immigrants from non-Western countries and their descendants’ is based on a ‘complex combination of criteria’, including country of birth, the nationality and country of birth of the parents and whether those countries of birth and/or nationality are considered, under Danish law, as ‘non-Western’ countries (§ 100).

On direct and indirect discrimination

A distinction based on ethnic origin constitutes direct discrimination if it involves a person being treated less favourable than another in a comparable situation (Art. 2(2)(a) RED). In line with the AG’s Opinion, the Court identifies two types of less favourable treatment that could result from the Danish legislation. The first is that tenants of public family housing in “transformation areas” face a higher chance of having their lease terminated compared to tenants with a similar lease in “vulnerable residential areas”, which are comparable to “transformation areas” in socioeconomic terms but where “non-Western immigrants and their descendants” do not form more than half of the population. It follows from the judgment that the risk of a lease being terminated early is, in itself, sufficient to constitute less favourable treatment, there is no requirement that this risk must have materialised. The Court moreover confirms what it had already decided in CHEZ, namely that there is direct discrimination when less favourable treatment results from a distinction based on ethnic origin, even if among those affected there are people who do not belong to the targeted ethnic group (§ 107). After all, tenants who are not themselves “non-Western” immigrants face the same risk of losing their homes as their “non-Western” neighbours. What is at stake here is a form of discrimination by association, although this concept is not mentioned explicitly by the Court.

Besides the risk of having their leases terminated, the Court notes a second form of harm that the applicants may have suffered, which is stigmatisation. Here the Court shows itself much more deferential to the Danish courts: it is up to the referring court to investigate if ‘the very name “transformation area”, which for the residents of areas classified as such gives rise to an increased risk of early termination of their lease, and which replaced the name ‘hard ghetto area’ is, at national level, offensive and stigmatising’ (§ 126). It is interesting that, at this point, the Court does show consciousness of the fact that the meaning of certain terms is socially constructed and can therefore differ over time and place, in this case leaving the Danish courts in a better position to determine the stigmatising connotations of the notion of “transformation areas”.

After having thus applied a broad definition of direct discrimination, which acknowledges both redistributive and recognitional harm, the Court moves on to examine the possibility of indirect discrimination. According to the Court itself, this analysis is necessary in case ‘the referring court concludes that the national legislation at issue […] does not constitute direct discrimination’ (§ 130). While this may at first sight seem a logical step to take, the Court’s consideration of indirect discrimination becomes less obvious if it is taken into account that the previous part of the judgment leaves very little to no room for the referring court to find that there has not been any direct discrimination. Indeed, once direct discrimination is established, the analysis can end there: discrimination is either direct or indirect, but not both, at least not on the same ground.

This move is troubling. The scheme at issue is plainly not one of indirect discrimination, and introducing this layer of analysis risks generating confusion rather than clarity for national courts. For one, if the referring court concludes that there is no direct discrimination because, after all, the criterion concerning “non-Western immigrants and their descendants” does not result in differential treatment based on ethnic origin, this would raise the question of what would then be the ethnic group (or groups) that could be put at a particular disadvantage, as required by Article 2(2)(b) RED. This is a question on which the judgment remains silent. The risk of confusion is further exacerbated by the fact that the Court’s own conceptualization of indirect discrimination is fraught with difficulties, as illustrated most clearly by its case law on religious symbols and headscarves (see here and here).

On the upside, the judgment does provide valuable clarification as to how the case should be assessed if viewed through the lens of indirect discrimination. This may prove practically significant, as experience shows that even where direct discrimination is formally removed, discriminatory effects often re-emerge—consciously or unconsciously—through ostensibly neutral criteria. One can easily imagine, for instance, that the Danish legislator might in the future abandon the explicit reference to “non-Western immigrants and their descendants”, while continuing to target specific neighbourhoods through socio-economic indicators that would, in practice, place that very group at a particular disadvantage. Some key takeaways from The Court’s analysis are: 1) that the aim of “ensuring successful integration of third-country nationals (TCNs)” is legitimate in principle; 2) the potential of the Danish scheme to actually promote social cohesion and integration is viewed as doubtful, not least because it does not apply to “vulnerable residential areas” that are socioeconomically comparable to “transformation areas” but without a majority “non-Western” population (§§ 162-163); 3) in addition to the requirements of appropriateness and necessity, which are expressly mentioned in Article 2(2)(b) RED, the Court examines the measures’ proportionality stricto sensu, something that has not commonly done before (see here and here); 4) regarding this proportionality stricto sensu, it must be taken into account that the right to respect for the home is a fundamental right protected by the EU Fundamental Rights Charter (Article 7) and that, according to ECtHR case law, the loss of one’s home constitutes “a most extreme form of interference” (§ 170).

Structural discrimination

In sum, there is a clear suggestion that the Danish scheme would also have to be considered as indirectly discriminatory – at least if a particular disadvantage to one or more ethnic groups can be established. The judgment can thus be readily qualified as an important blow to Denmark’s “ghetto” policy. Still, from a perspective of structural equality several problematic aspects remain – we round off this blog by mentioning two of them. The first, already raised by one of us elsewhere, is that the judgment entrenches the existing distinction between EU citizens and third-country nationals whereby only the latter are deemed in need of integration. This becomes evident in the indirect discrimination analysis, where it is pointed out that “integration” is a legitimate public interest only where third-country nationals are concerned (§ 151) – the subtext being that EU citizens are by definition already integrated.

Second, the Court seems to leave open the possibility that, under certain circumstances, “gentrification” measures such as those imposed by Denmark might be proportionate. This suggestion is troubling, given that those measures reflect a discriminatory logic that runs even deeper than the use of stereotypes in legislative documents or the harm suffered by individual tenants who are forced to leave their homes. At the basis of Denmark’s (or perhaps any) integration policy lies the rationality of Goldberg’s ‘racial state’ – the notion that at its core the modern state serves to protect public order by excluding the racialised other. It follows that any proportionality analysis cannot meaningfully abstract from the fact that the policy is premised on the stigmatization of racialised and impoverished communities. Yet is seems that the Court failed to recognise this, unlike in CHEZ where it expressly acknowledged the offensive and stigmatising nature of the practice at issue (§§84 and 108).

In short, there are several positive points to be noted about the judgment. These are, in particular, the Court’s acknowledgement that race discrimination does not turn on the presence of a homogeneous ethnic group, its recognition of the stigmatisation inherent in the so-called ‘ghetto law’ and its engagement with international instruments and proportionality in the indirect discrimination analysis. Still, the Court stops short of addressing the central issue at stake: the systemic and structural character of racialised exclusion. By avoiding a direct confrontation with structural racism, the Court ultimately does not get to the bottom of what makes the Danish ‘ghetto law’ so deeply problematic.

The post The EU must solve the credibility deficit in its enlargement policy first appeared on LSE European Politics.

The post Could artificial intelligence advance European federalism? first appeared on LSE European Politics.

Ana Luísa Azevedo Lopes (master’s student in European Union Law at the School of Law of the University of Minho)

1. Introduction

The protection of animal welfare has become an increasingly visible element of European Union (“EU”) action, particularly in policy areas traditionally dominated by economic considerations. The signing in January 2026 of the EU–Mercosur Free Trade Agreement (“EMTA”)[1] has made this a particularly salient issue for animal rights organisations and European citizens, serving as a contemporary example of how animal welfare is addressed both internally, and in the Union’s external action. Late in February 2026, President Ursula Von der Leyen announced that the Commission is moving forward with the provisional implementation of the agreement, although she stated that it can only be fully concluded after consent from the European Parliament.[2] This decision makes the subject of this article even more timely, as we shall examine the potential effects of such agreement on animal welfare matters across the EU.

That being said, whilst EU legislation has long addressed animal welfare in sectors such as agriculture, transport, and slaughter, the introduction of Article 13 Treaty on the Functioning of the European Union (“TFEU”) marked a significant symbolic development at the level of EU primary law, formally recognising animals as sentient beings.^[3] Despite this recognition, the legal status and practical relevance of Article 13 TFEU remain uncertain. The provision does not establish concrete obligations or enforceable rights but rather requires that animal welfare be taken into account in the definition and implementation of certain Union policies. This raises a fundamental question as to how animal welfare operates within the EU legal order when it comes into tension with other objectives pursued by the Union, particularly those linked to economic integration and trade.^[4]

This paper approaches Article 13 TFEU through the lens of normative conflict, examining how animal welfare is balanced against competing policy objectives within the EU legal framework. Particular attention is paid to the case law of the Court of Justice of the European Union (“CJEU”), focusing on the Zuchtvieh-Export judgment (C-424/13),^[5] which offers a revealing illustration of the limits attributed to animal welfare considerations in contexts involving cross-border economic activity and trade with third countries. The choice of this judgment is justified by its clear exposure of the tension between the protection of animals and the logic of market integration. In addition, the analysis incorporates the Centraal Israëlitisch Consistorie van België judgment (C-336/19),^[6] which provides a critical perspective on the balancing of animal welfare against fundamental rights, specifically the freedom of religion. This case is essential to the study, as it demonstrates the Court’s evolving interest, while simultaneously exposing the structural challenges of integrating Article 13 TFEU into the Union’s complex hierarchy of values.

This analysis serves to highlight the broader structural challenges faced by Article 13 TFEU when animal welfare considerations intersect with the Union’s commercial and strategic interests.

2. The recognition of animal welfare in EU primary law

The recognition of animals as “sentient beings” constitutes a significant rupture with the traditionally anthropocentric view of law, in which animals were regarded primarily as economic assets or factors of production.

The first shift emerged with Protocol No. 10 annexed to the Treaty of Amsterdam (1997), which introduced for the first time an obligation for the Union and the Member States to pay full regard to the welfare requirements of animals when formulating policies on agriculture, transport, the internal market, and research. Although protocols annexed to the Treaties possess the same legal value, as said in Article 51 of the Treaty on the European Union (“TEU”), constituting primary law, their positioning outside the main body of the Treaties has been argued to contribute to their reduced visibility and limited normative influence in practice. In this sense, despite representing an important legal recognition of animal welfare, it wasn’t until the Treaty of Lisbon (2009) that this issue gained a spotlight position, as outlined in Article 13 of the TFEU.^[7]

This article serves as a cross-cutting principle, whose primary function is to integrate animal welfare concerns into policies that pursue diverse objectives. The creation of this article represents a milestone in the evolution of EU law regarding animal protection. By clearly recognising animals as “sentient beings”, EU primary law moves away from a strictly instrumental conception of animal protection, traditionally associated with economic efficiency or product quality.^[8]

This process of recognising the animals’ capacity to feel is part of a broader movement toward affirming ethical values within the Union’s legal order, in line with the increasing importance attributed to environmental protection, public health, and fundamental rights. With Article 13 TFEU recognising animals as “sentient beings”, the EU legally acknowledges that animals have the capacity to feel pain, suffering, and pleasure.^[9] It functions as a horizontal clause that mandates animal welfare be considered transversally across various areas of Union competence, namely in the definition and implementation of policies on agriculture, transport, the internal market, and research.^[10]

This transversality confers a significant legal and political character upon Article 13 TFEU, insofar as it obliges European legislators and decision-makers to consider animal welfare within the decision-making process. Article 13 TFEU has been commonly characterised as a horizontal clause or integration principle, comparable to other value-oriented provisions contained in Articles 8 to 12 TFEU. Rather than establishing autonomous regulatory obligations, such provisions operate as interpretative and coordination mechanisms designed to ensure that certain ethical and social values are systematically integrated into the formulation and implementation of Union policies. In this sense, Article 13 TFEU can be understood as a programmatic norm whose legal relevance materialises primarily through legislative development and judicial interpretations. Its normative force, therefore, depends on the balancing exercise carried out between animal welfare and other competing Union objectives, particularly those connected to economic integration and market functioning.^[11]

Article 13 TFEU must be interpreted in light of the overarching principles of EU law, particularly the principle of consistency and the principle of proportionality.^[12] As a provision of primary law, it functions as a mandatory parameter for the legality of Union acts. This implies that the protection of animals, as sentient beings, has been integrated into the Union’s “constitutional” core, requiring a systematic dialogue with other foundational objectives, such as the promotion of the Internal Market and the Common Agricultural Policy. Consequently, any legislative or administrative measure that affects animal welfare must undergo a rigorous balancing test. In this context, the principle of sincere cooperation also plays a role, as it obliges both the Union and Member States to ensure that the effectiveness of Article 13 TFEU is not undermined by conflicting secondary legislation or external trade commitments that ignore the ethical boundaries established by the treaties.^[13]

However, it is important to underline that Article 13 TFEU does not establish concrete material obligations, nor does it create directly enforceable subjective rights (rights with direct effect). Its primary function is advisory and guiding, conditioning the decision-making process without imposing specific results. In other words, this duty of care is not equivalent to an obligation of result, remaining largely dependent on political choices and proportionality assessments. Still, the “spotlight position” granted by the Treaty of Lisbon does not just change the status of the animal, it fundamentally recalibrates the hierarchy of European values, forcing a move away from anthropocentric dominance toward a more integrated, ethical legal order.^[14]

3. The structural limits of Article 13 TFEU and its position within the EU legal order

Although Article 13 TFEU has elevated animal welfare to the status of primary law, its effectiveness faces significant obstacles that limit its scope as a norm of absolute protection.

One of the main weaknesses of this Article lies in its lack of direct effect. The provision is not sufficiently clear, precise, or unconditional to be invoked autonomously by individuals before national or European courts. Consequently, its legal impact depends on the mediation of the Union legislator and the interpretation provided by the CJEU. Thus, Article 13 TFEU functions merely as an interpretative aid, a factor to be considered in a decision, rather than a legal basis capable of leading to a favourable ruling for animal welfare.

Furthermore, Article 13 TFEU includes a safeguard clause requiring respect for the “religious rites, cultural traditions and regional heritage” of Member States. This provision creates an implicit hierarchy where animal welfare, despite being recognised, yields to freedom of religion – Article 10 of the Charter of Fundamental Rights of the European Union (“Charter”) –, or the cultural identity of the States. This significantly weakens the scope of the norm by introducing a broad exception that allows animal welfare to be relativised in the name of other values.^[15]

Take as an example, case C-336/19, Centraal Israëlitisch Consistorie van Bergië and Others, which addressed the compatibility between national legislation requiring prior stunning of animals before slaughter and the protection of freedom of religion. In this judgment, the Court recognised that animal welfare constitutes a legitimate objective of general interest within the EU, supported by Article 13 TFEU. However, it also held that such protection must be balanced against the fundamental right to freedom of religion based in Article 10 of the Charter.

The court ultimately concluded that Member States may impose mandatory stunning requirements provided that such measures respect the principle of proportionality and do not unduly restrict religious practices. This decision illustrates the structural position of animal welfare within the Union legal order: although recognised as a constitutional value, it operates within a balancing framework and does not automatically prevail over other fundamental rights.^[16]

These limitations are reflected in the position occupied by animal welfare within the EU’s hierarchy of values. Although recognised as a value, animal welfare often appears subordinated to economic objectives, the functioning of the internal market, and competitiveness. This subordination becomes particularly evident in situations of conflict between Union policies, where Article 13 TFEU functions more as a balancing principle than as a true normative limit.

4. Article 13 TFEU in the case law of the Court of Justice of the European Union

The case law of the CJEU has been consolidating an interpretation of Article 13 TFEU which, despite its seemingly restrictive nature as a framework provision, reveals an increasing hermeneutic force. The Court recognises animal welfare as a legitimate and cross-cutting interest of the Union, yet its practical application oscillates between contextual integration and the assignment of a decisive role in resolving complex disputes.

Taken together, the case law of the CJEU reveals a consistent interpretative pattern in which animal welfare is recognised as a legitimate Union value, but rarely operates as an overriding principle, instead being integrated through proportionality-based balancing exercises involving competing economic interests, policy objectives, and fundamental rights.

The ambivalent nature of this approach becomes particularly evident when comparing the Court’s reasoning across different areas of Union action. In case C-336/19, Centraal Israëlitisch Consistorie van België and Others, as mentioned above, the Court was called upon to assess whether national legislation requiring prior stunning of animals before slaughter was compatible with the protection of freedom of religion. In this judgment, the Court expressly acknowledged that the protection of animal welfare constitutes an objective of general interest within the Union’s legal order and reflects evolving ethical standards in European societies, as recognised by Article 13 TFEU.^[17] However, the Court simultaneously emphasised that such protection must be compatible with the fundamental right to freedom of religion, guaranteed by Article 10 of the Charter. By accepting that Member States may impose stunning requirements subject to proportionality review, the Court confirmed that animal welfare, despite its constitutional recognition, remains structurally positioned within a framework of normative balancing rather than functioning as an absolute legal constraint.^[18]

The Zuchtvieh-Export judgment (C-424/13) constitutes the paradigmatic example of this ambivalent approach in the field of economic activity and external trade. Called upon to rule on the application of Regulation No. 1/2005 concerning the transport of live animals during long-distance journeys to third countries, the Court adopted an interpretation strongly influenced by the normative orientation of Article 13 TFEU.^[19] Departing from a merely programmatic understanding of the provision, the Court established that Article 13 TFEU imposes a rigorous procedural obligation on national authorities responsible for authorising animal transport.

By ruling that animal welfare must be guaranteed from the place of departure to the final destination – including locations outside the EU territory –, the CJEU effectively extended the territorial reach of Union animal welfare standards. In doing so, the Court transformed the obligation to pay “full regard” to animal welfare into a prerequisite for the legality of administrative authorisation decisions.^[20] The practical consequence of this interpretation is that the export of live animals becomes conditional upon compliance with Union welfare standards even in third countries, thereby reinforcing the ethical dimension of internal market regulation.

Nevertheless, the decision did not elevate Article 13 TFEU to an autonomous legal basis capable of independently determining the outcome of the dispute. Instead, the Court preserved the structural priority of market integration and commercial viability while simultaneously requiring that economic operators demonstrate, through realistic and detailed journey planning (or journey log), that animal sentience will be adequately protected. Even while admitting limitations to animal protection for the sake of commercial viability, the Court shifts the burden of proof: it is up to the economic operator to demonstrate that animal sentience will be respected.^[21]

When read together, these judgments demonstrate that Article 13 TFEU functions as a principle of normative integration and judicial filtering. It does not prevent the subordination of animal welfare to other policy objectives, particularly those related to the Common Agricultural Policy and the Common Commercial Policy. However, it requires that any limitation on animal protection must be objectively justified, proportionate, and compatible with the ethical commitments formally recognised within the Union’s constitutional framework.

5. Animal welfare and EU trade policy: the EMTA as a case study

Article 13 TFEU represents a significant development in the constitutional evolution of EU law, as it expressly recognises animals as sentient beings and imposes on the EU institutions and the Member States the obligation to pay full regard to animal welfare requirements when formulating and implementing certain policies. This recognition has often been presented as a decisive step in the process of the “constitutionalisation” of animal protection within the Union. However, an analysis of legislative practice and, in particular, of the EU’s external action demonstrates that the effectiveness of this provision remains limited, revealing a persistent tension between the protection of animal welfare and other objectives regarded as priorities, such as the functioning of the internal market and the Common Commercial Policy.^[22] The EMTA constitutes a paradigmatic example of this tension and provides a useful lens through which to assess whether Article 13 TFEU operates as a genuinely effective norm or rather as a largely symbolic one.

The EMTA aims to establish one of the largest free trade areas in the world by progressively reducing tariffs and facilitating trade flows between the two blocs. From the perspective of the EU, the agreement is presented as an instrument designed to strengthen geopolitical partnerships, expand export opportunities for European industries, and enhance economic competitiveness in a globalised market. Among its most significant commercial effects is the expansion of access to the EU market for agricultural products originating in Mercosur countries, particularly beef, poultry, and other products of animal origin. While these developments may generate economic benefit and increase consumer choice within the Union, they simultaneously raise complex regulatory and ethical concerns in relation to animal welfare.^[23]

The EU has progressively developed a comprehensive regulatory framework governing animal welfare standards applicable to its internal production systems, including rules relating to farming conditions,^[24] transport,^[25] and slaughter practices.^[26] These regulatory developments partly reflect the normative orientation established by Article 13 TFEU and demonstrate the increasing integration of ethical considerations into Union agricultural and food production policies.^[27] However, the EMTA introduces a potential regulatory asymmetry insofar as it allows increased importation of animal-derived products originating from production systems that may not operate under standards fully equivalent to those imposed to Union producers.

It should be considered that the conclusion of the EMTA does not automatically result in a lowering of internal Union standards. Products entering the European market remain subject to compliance with existing EU sanitary and regulatory requirements. These mechanisms operate as safeguards intended to prevent the entry of products that fail to meet minimum regulatory thresholds.^[28] However, the effectiveness of these safeguards depends significantly on monitoring procedures, certification systems, and enforcement capacity. Potential regulatory inconsistencies or failures in compliance control may ultimately give rise to disputes requiring judicial review before the CJEU, thereby illustrating the indirect and reactive nature of existing enforcement mechanisms.

Furthermore, although the EMTA contains chapters dedicated to sustainable development and establishes institutional mechanisms promoting regulatory cooperation, exchange of technical expertise, and policy dialogue between the parties, these provisions remain largely framed within soft-law structures. The agreement does not introduce binding conditionality mechanisms specifically designed to ensure compliance with high animal welfare standards by third-country producers. As a result, the normative influence of Article 13 TFEU within the context of the Union’s external trade policy remains structurally limited.^[29]

Recent exploratory studies and policy analyses have sought to evaluate the potential impact of the EMTA on animal welfare outcomes. These studies suggest that increased demand for animal-derived products may incentivise the expansion of intensive livestock production systems within Mercosur countries, potentially resulting in increased animal suffering and associated environmental impacts, including deforestation and biodiversity loss. Although the empirical projections surrounding these impacts remain subject to ongoing academic policy debates, they reinforce concerns regarding the capacity of current trade instruments to effectively integrate animal welfare considerations into global supply chains.^[30]

The EMTA, therefore, illustrates the broader structural challenges faced by Article 13 TFEU when animal welfare considerations intersect with the Union’s Common Commercial Policy, governed primarily by Article 207 TFEU. While animal welfare is formally recognised as a constitutional value within the Union legal order, the architecture of international trade policy continues to be driven essentially by economic and market-related goals. Within this framework, animal welfare considerations tend to operate as a secondary normative factor rather than decisive regulatory constraints. Consequently, the EMTA highlights the dual character of Article 13 TFEU within the Union’s external action. The provision contributes to shaping political discourse and regulatory expectations surrounding animal protection but lacks the binding enforcement mechanisms necessary to ensure full compliance with international trade relations. Although the sustainable provision is a very important step, there is no way to enforce EU standards in third countries.

Criticism voiced by organisations such as Eurogroup for Animals emphasises that the agreement may have devastating impacts on millions of animals by encouraging the expansion of intensive livestock farming in Mercosur countries, often associated with practices that would be unlawful under EU law.^[31] Beyond the direct suffering inflicted on animals, there are significant indirect effects, including increased deforestation and environmental degradation, which further exacerbate the inconsistency between the values proclaimed by the European Union and its external action. In this way, the EU maintains high animal welfare standards within its own territory while simultaneously ending up accepting, through trade, products derived from production systems that undermine those very values, effectively externalising animal suffering beyond its borders.

Despite the constitutional recognition of animal sentience, the EU has been unable to ensure that this recognition produces tangible legal consequences when confronted with other policies regarded as strategic, such as the Internal Market and the Common Commercial Policy. Animal welfare thus emerges as a subordinated value, frequently invoked in legal and political discourse but readily sacrificed when it conflicts with economic interests.^[32] In this sense, the EMTA reinforces the view that Article 13 TFEU oscillates between symbolism and limited effectiveness, functioning more as a guiding principle than as a genuine substantive constraint on the European Union’s conduct in international trade.^[33]

However, it is important to note that the Union’s current legislative agenda appears to be testing a response to these asymmetries. The “Omnibus Package”, part of the Farm to Fork strategy, aims to revise EU animal welfare legislation to align production standards with the latest scientific evidence and citizens’ ethical demands. In the context of external relations, this legislative package is particularly relevant as it reopens the debate on the application of reciprocity clauses (or mirror clauses). It seeks to ensure that products imported from third countries, such as the Mercosur bloc, do not undermine the sustainability and animal protection standards imposed on European producers. It remains to be seen whether these legislative intentions will possess sufficient legal strength to overcome the primacy of World Trade Organization free trade rules, which have historically constrained the external effectiveness of Article 13 TFEU.^[34]

6. Conclusion

The analysis developed throughout this study allows for the conclusion that Article 13 TFEU constitutes a fundamental axiological and symbolic milestone, elevating animal welfare to the status of a transversal value of the European Union and providing legal recognition of animal sentience. However, the transition of this “principle-norm” toward full legal effectiveness remains faltering, particularly when confronted with Union policies endowed with consolidated normative and economic weight, such as the Common Agricultural Policy and the Common Commercial Policy.^[35]

Article 13 TFEU should not be viewed as a provision devoid of relevance; yet its current function is predominantly guiding and subordinate. As demonstrated by the case law of the CJEU and, most flagrantly, by the negotiations of the EMTA, animal welfare tends to be sacrificed at the altar of international competitiveness and trade liberalisation. The agreement exposes a legal incoherence: while the Union imposes high ethical standards on its internal producers, it tolerates the importation of products from third countries that do not observe equivalent requirements, resulting in an undesirable “outsourcing of cruelty”.^[36]

This reality raises critical questions regarding the coherence of the European project and the actual standing of animal welfare within the Union’s hierarchy of values. For Article 13 TFEU to cease being an exercise in constitutional rhetoric and become a true normative limit on the EU’s actions, the adoption of equivalent standard requirements and strict conditionality mechanisms in external relations is imperative. Without a robust integration that binds trading partners to European ethical standards, Article 13 TFEU will remain, to a large extent, a symbolic affirmation of sentience, lacking the imperative force necessary to shape the action of the EU as a global actor committed to animal ethics.

[1] European Commission, “EU and Mercosur sign historic agreement creating one of the largest free trade zones in the world”, January 20, 2026, accessed February 16, 2026, https://ec.europa.eu/commission/presscorner/detail/en/ac_26_163.

[2] Peggy Corlin and Vincenzo Genovese, “Von der Leyen to implement contentious Mercosur trade deal despite MEPs’ legal challenge”, Euronews, 27 February 2026, https://www.euronews.com/my-europe/2026/02/27/commission-to-implement-contentious-mercosur-trade-deal-despite-eu-parliament-opposition.

[3] Justine Coudron, “The impact of Article 13 TFEU on the free movement of animals and animal products” (Master’s diss., Ghent University, 2023), 40–43.

[4] Émilie Delcher, “Introductory observations on animal welfare: a concept applicable to wild animals in European Union legislation?”, trans. Heron Gordilho and Lyliam Botteau, Brazilian Journal of Animal Law, v. 17, no. 1 (2022): 1–19.

[5] Judgment CJEU Zuchtvieh-Export, 23 April 2015, Case C-424/13, ECLI:EU:C:2015:259, recital 2.

[6] Judgment CJEU Centraal Israëlitisch Consistorie van België, 17 December 2020, Case C-336/19, ECLI:EU:C:2020:1031, recitals 1-2.

[7] José Martinez and Cara von Nolting, “Review: ‘animal welfare’ – a European concept”, Animal, v. 17 (2023): 3, doi: 10.1016/j.animal.2023.100839.

[8] Diane Ryland, “Taking stock of Art. 13 TFEU in EU agriculture: reading Art. 13 as a whole,” European Papers, v. 8, no. 1 (2023): 192-196, doi: 10.15166/2499-8249/646.

[9] Heather Browning and Jonathan Birch, “Animal sentience”, Philosophy Compass, v. 17, no. 5 (2022): 1-2, doi: 10.1111/phc3.12822.

[10] Evangelia Psychogiopoulou, “The horizontal clauses of Arts 8-13 TFEU through the lens of the Court of Justice”, European Papers, v. 7, no. 3 (2022): 1358-1363, doi: 10.15166/2499-8249/618.

[11] Evangelia Psychogiopoulou, “Unravelling the complexities of the horizontal clauses of Arts 8-13 TFEU: an explanation of the special section”, European Papers, v. 8, no. 1 (2023): 221-226, doi: 10.15166/2499-8249/647.

[12] Tor-Inge Harbo, “The function of the proportionality principle in EU Law”, European Law Journal, v. 16, no. 2 (2010): 164-165.

[13] Jurian Langer and Wolf Sauter, “The consistency requirement in EU Law”, Columbia Journal of European Law, v. 24, no. 1 (2017): 43-45.

[14] Delcher, “Animal welfare”, 6–8.

[15] Ilja Richard Pavone, “Towards an EU animal welfare law: the case of animal testing and the limits of new welfarism”, Animal & Natural Resources Law Review, v. 16 (2020): 208–212.

[16] Judgment CJEU Centraal Israëlitisch Consistorie van België, recitals 81 and 90.

[17] Judgment CJEU Centraal Israëlitisch Consistorie van België, recitals 25-32.

[18] Judgment CJEU Centraal Israëlitisch Consistorie van België, recitals 47-56.

[19] Judgment CJEU Zuchtvieh-Export, 23 April 2015, Case C-424/13, ECLI:EU:C:2015:259, recitals 1-2.

[20] Judgment CJEU Zuchtvieh-Export, recitals 20-22.

[21] Judgment CJEU Zuchtvieh-Export, recitals 25-30.

[22] As previously seen with the analysis of EU judgments C-336/19 and C-424/13.

[23] Council of the European Union, “Council decision on the signing and provisional application of the interim agreement on trade between the European Union, of the one part, and the Southern Common Market, the Argentine Republic, the Federative Republic of Brazil, the Republic of Paraguay and the Oriental Republic of Uruguay, of the other part,” 12417/1/25 REV 1, 8 January 2026.

[24] Council Directive 98/58/EC of 20 July 1998 concerning the protection of animals kept for farming purposes.

[25] Council Regulation (EC) No 1/2005 of 22 December 2004 on the protection of animals during transport and related operations.

[26] Council Regulation (EC) No 1099/2009 of 24 September 2009 on the protection of animals at the time of killing.

[27] European Union, Council Regulation (EC) No 1099/2009 of 24 September 2009 on the protection of animals at the time of killing, Official Journal of the European Union L 303 (18 November 2009);
European Union, Council Regulation (EC) No 1/2005 of 22 December 2004 on the protection of animals during transport and related operations, Official Journal of the European Union L 3 (5 January 2005).

[28] Leonardo Fabio Pastorino and Washington Carlos de Almeida, “Review: impact of bilateral trade on the promotion of animal welfare rules. The case of trade relations between the European Union and Mercosur”, Animal, v. 17 (2023): 1-4, doi: 10.1016/j.animal.2023.100837.

[29] As discussed by the EU Intergroup on the Welfare and Conservation of Animals, specifically by Daniel Pérez Vega – project officer and animal welfare at Eurogroup for Animals: “Another problem in the agreement is that all the sustainability provisions in the agreement, including all the texts that were published in December, cannot offset the negative consequences of the agreement. Because all the sustainability provisions, they are very nice, there’s very nice wording in them, there’s no way to enforce them. And we know this because we have the same problem. I sit in most of the other and monitor the implementation of agreements, and none of them have had a cooperation action activated, because most of the time you need the political willingness of partners to cooperate, and there’s no obligation to do so”. See EU Intergroup on the Welfare and Conservation of Animals, “EU-Mercosur: why the trade deal still fails animals”, June 19, 2025, accessed February 16, 2026, https://www.animalwelfareintergroup.eu/calendar/eu-mercosur-why-trade-deal-still-fails-animals.

[30] Policy analysis promoted by animal welfare organizations suggest that increased EU’s demand for animal-derived products may encourage the expansion of intensive farming systems in Mercosur countries. These reports emphasise that export-oriented production is typically dominated by large industrial producers rather than small-scale farmers, potentially increasing animal suffering and creating a mismatch between EU consumer expectations and production realities abroad. See Eurogroup for Animals, “The EU-Mercosur.”

[31] Eurogroup for Animals, “EU-Mercosur.”

[32] Council of the European Union, “Council decision on the signing, on behalf of the Union, and on the provisional application of the partnership agreement between the European Union and its Member States, of the one part, and the Southern Common Market, the Argentine Republic, the Federative Republic of Brazil, the Republic of Paraguay and the Oriental Republic of Uruguay, of the other part.”

[33] Eurogroup for Animals, “EU-Mercosur.”

[34] European Commission, “Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions: a farm to fork strategy for a fair, healthy and environmentally-friendly food system”, COM(2020) 381 final, Brussels, May 20, 2020, 4–10.

[35] Ryland, “Taking stock”, 219.

[36] Eurogroup for Animals, “EU-Mercosur”.

Picture credit: by Matthias Zomer on pexels.com.

The post Emily Jackson: “We’ve gone from thinking the world is massively overpopulated to worrying about falling birth rates” first appeared on LSE European Politics.

The post The politics of belonging – why enlargement is now a strategic priority for the EU first appeared on LSE European Politics.

The post Is Donald Trump right about free-riding in NATO? first appeared on LSE European Politics.