The Council of the EU’s Position on the Return Proposal: Trump-Inspired Approaches Dominate

 

Dr. Izabella Majcher, Independent Consultant

Photo credit: Alamy stock photo

 

On 11 March 2025, the European Commission published a proposal for a Return Regulation intended to replace the 2008 Return Directive. The stated objective of the proposal is to increase the number of people returned, with the Commission regretting that only around 20% of persons issued with a return decision actually leave. As the proposal explains, it seeks to streamline and harmonise return procedures, attributing the perceived ineffectiveness of EU return policy to divergent national approaches. In this spirit, the Commission proposes replacing the directive with a regulation that would establish “modern, simplified and common” procedures. EU institutions have repeatedly asserted that the current return framework is complex and outdated. However, this claim is difficult to verify, as is the assumption that the proposed regulation would effectively increase return rates. This is largely due to the Commission’s failure to conduct a prior impact assessment, despite this being required under its own Better Regulation Guidelines.

On the side of the European Parliament, the draft report on the Commission’s proposal was presented to the Committee on Civil Liberties, Justice and Home Affairs (LIBE) on 11 November 2025, and the Parliament is currently working towards adopting its final position. The other co-legislator, the Council of the European Union, adopted its “general approach” on 8 December 2025, during the Justice and Home Affairs (JHA) Council, alongside its position on the safe third country concept and on a common EU list of safe countries of origin. This JHA Council was the final meeting under Denmark’s Presidency. Denmark—widely known for its restrictive asylum and migration policies—sought to secure Member States’ agreement on these files before the end of its term (for instance, the Council removes the future mandatory character of the mutual recognition of return decision proposed by the Commission, on which the agreement among Member States proved difficult).

While the Commission’s proposal already displays an overall coercive approach and restricts human rights safeguards—as highlighted in a joint statement by over 200 civil society organisations—the Council’s position further erodes existing protections. This non-exhaustive analysis examines ten key measures introduced or reinforced by the Council that are particularly problematic from a human rights perspective.

 

1) Return decision and the risk of refoulement: the Council explicitly allows disregard of the principle of non-refoulement. It provides that authorities should not be precluded from issuing a return decision where there is a risk of refoulement (Art. 7(4a)). In such cases, the return decision should state that removal to the country concerned is postponed under Art. 14(1)(a). Given that a return decision may be enforceable, issuing it despite an identified risk of refoulement engages the State’s responsibility under Article 3 ECHR. Issuing a return decision in such circumstances—irrespective of a possible postponement of removal—may lead to refoulement in practice, for example if the postponement decision is not adopted in time. Moreover, as the suspensive effect of appeals is significantly curtailed (see below (2)), a person may be removed before such a decision is taken. This approach is also questionable in terms of effectiveness: postponement procedures would still need to be carried out, while the person would remain in an irregular situation, thereby undermining the stated objective of increasing return rates.

In practice, this provision appears to facilitate transfers to third countries acting as “return hubs” (see below (9)) during the postponement of removal to the country designated in the return decision. However, the prohibition of refoulement also covers indirect (chain) refoulement. Transfer to an intermediary country does not absolve the sending State of its responsibility under Article 3 ECHR to ensure that the “return hub” offers sufficient guarantees against onward removal to a risk of refoulement.

2) Procedural safeguards and the right to an effective remedy: the proposal further increases the risk of refoulement by weakening procedural safeguards to challenge return decisions. As introduced by the Commission and reinforced by the Council, a return decision need not specify the country of return (Art. 7(4)), while a removal decision may identify more than one country (Art. 4(4b)). This undermines legal certainty and clarity and places an excessive burden on individuals, who may be required to challenge removal in relation to several potential destinations. The proposal foresees a non-refoulement assessment prior to removal, which is particularly important given that a return decision may be issued despite an identified risk of refoulement (see above (1)). Accordingly, where there are indicators that removal would breach the principle of non-refoulement, the person should be referred to an appropriate procedure, including the asylum procedure, or the risk should be assessed in accordance with national law. However, this assessment may be carried out by the same authorities that issued or enforce the return decision, and may be dispensed with entirely where a prior assessment exists or where the relevant information is deemed insufficiently substantiated (Art. 12(3)–(3a)).

Remedies are also weakened. The Council deletes the Commission’s proposal to suspend enforcement of a return decision where there is a risk of refoulement. Automatic suspensive effect is no longer guaranteed; instead, Member States may grant it only upon request (Art. 28(1)). Given the irreversible nature of the harm, Art. 13 ECHR requires automatic suspensive effect where return is challenged on non-refoulement grounds.

3) Grounds for non-return: under the current Return Directive, Member States may grant an autonomous residence permit or other authorisation offering a right to stay for humanitarian, compassionate or other reasons. In such cases, no return decision shall be issued, or an existing decision must be withdrawn or suspended for the duration of the authorisation (Art. 6(4), current Directive). While the Commission retains this possibility (Art. 7(9), Commission proposal)—albeit outside the exceptions to the obligation to issue a return decision in Art. 8—the Council removes it entirely. Its position contains no human-rights-based exception to the obligation to issue a return decision. Yet such exceptions are essential to ensure respect for rights that may prevent removal, including non-refoulement, the right to private and family life, or other circumstances rendering return unlawful or inappropriate, such as serious health conditions. Eliminating the possibility to grant a residence permit in such situations risks violations of international and EU human rights obligations and conflicts with Member States’ own legal frameworks, many of which provide for regularisation mechanisms. Not every person in an irregular situation can be removed. Issuing return decisions in cases where removal would violate human rights is likely to generate litigation and leave individuals in legal limbo—ultimately undermining, rather than enhancing, the effectiveness of return policy.

4) Detention: the Commission’s proposal significantly expands the legal basis for detention. In addition to the two grounds provided under the Return Directive—risk of absconding and hampering return—it introduces three new grounds: “security risks” (see below (8)), determination or verification of identity or nationality, and non-compliance with restrictive measures (see below (5)) (Art. 29(3)). As discussed elsewhere, these grounds are broadly framed and raise serious concerns as to their necessity and justification. The Council further extends this list by allowing detention on “other relevant, necessary and proportionate grounds” to ensure an effective return procedure, as defined in national law. This open-ended clause undermines legal certainty and permits unchecked and potentially arbitrary detention, in violation of the right to liberty.

The Commission also introduces extensive criteria for assessing the risk of absconding, including three criteria establishing a rebuttable presumption and eight criteria for an overall assessment (Art. 30, Commission proposal). The Council expands both lists further, adding undefined criteria under national law (Art. 21a, Council version). This disproportionate approach risks categorising nearly every person in an irregular situation as a potential absconder, thereby transforming immigration detention from an exceptional measure of last resort into a standard tool of migration control, contrary to international human rights standards.

Despite the requirement under international law that immigration detention be as short as possible, the Commission extends the maximum period of detention from 18 months (6 + 12) to 24 months (12 + 12) (Art. 32(3), Commission proposal). This detention period is exceptionally long and unnecessary, particularly given that immigration detention concerns persons who have not committed a criminal offence and who do not benefit from safeguards applicable under criminal law. The Council allows for repeated six-month extensions of the detention period where there is a risk of absconding and a reasonable prospect of removal arises from specific changes in circumstances (Article 32(3a), Council version). Such potentially indefinite detention amounts to arbitrary detention and violates the right to liberty. In addition, the Council clarifies that these detention limits apply per Member State, allowing a new period of detention to be imposed if the person moves to another EU country (Art. 32(3)).

Procedural safeguards are further weakened: the Council reduces the frequency of detention reviews from three to six months and removes the 15-day deadline for an initial judicial review (Art. 33).

5) Cooperation duties and restrictive measures: the Commission’s proposal imposes extensive cooperation obligations and restrictive measures on persons subject to return, which the Council substantially expands. The resulting web of detailed—and at times overlapping—obligations undermines legal clarity and proportionality. The obligation to cooperate (Art. 21) encompasses fifteen categories of obligations, including remaining available to authorities (such as appearing before third-country authorities for the issuance of travel documents or verification of identity or nationality), providing extensive information, and complying with third-country entry requirements, including health-related conditions. The list is open-ended, as it refers to national law. In practice, many of these obligations may be difficult or impossible for individuals to fulfil.

Non-compliance with cooperation duties triggers the mandatory imposition of restrictive measures, including residence requirements, reporting obligations, and other measures under national law (Art. 23(1)). It may also lead to further sanctions, such as refusal of allowances, extension of entry bans up to 20 years, criminal sanctions including imprisonment, or other penalties provided for under national law (Art. 23b(1)). Failure to cooperate also constitutes a criterion for establishing a risk of absconding (Art. 21a(2)(g)), which may in turn justify the imposition of so-called “alternatives to detention for preventing absconding”—including electronic monitoring and measures defined in national law (Art. 23c(1))—as well as detention itself (Art. 29(3)(a)). Moreover, failure to comply with specific cooperation duties is an independent ground for detention (Art. 29(3)(e)).

In practice, virtually any person subject to return may fail to comply with one or more of these extensive obligations and thus become exposed to a cascading set of restrictive measures and sanctions. The breadth of these measures fails to meet the principles of necessity and proportionality and leaves significant scope for arbitrary application.

6) “Investigative measures”: the Council introduces “investigative measures” (Art. 23a), allowing authorities—without the person’s consent—to search persons subject to return, their homes and other relevant premises, and to seize personal belongings, including electronic devices. Such measures may also be further defined under national law. While safeguards are foreseen, such as necessity, proportionality, respect for fundamental rights and access to remedies, the breadth of this provision raises serious concerns. In practice, it would permit police to raid homes, shelters and other premises that accommodate or support persons in an irregular situation, with significant implications for the right to privacy.

7) Entry ban: while the Commission already expands the use and duration of entry bans, the Council amplifies this approach further. The Commission introduces the possibility of imposing an entry ban without a prior return decision where an irregular stay is detected during exit border checks. The Council removes the Commission’s safeguards, namely the requirements that such a ban be justified by the individual circumstances of the case, respect the principle of proportionality, and safeguard the rights of defence. The Council’s position effectively eliminates the rights of defence by allowing an entry ban to be imposed and notified after the person has already left the territory (Art. 10(4)). It further allows the imposition of an entry ban without a prior return decision where a person departs before such a decision is issued (Art. 10(4a)). This raises serious concerns as to how authorities can assume, without any prior assessment, that a person in an irregular situation would not have a right to stay. Crucially, this deprives individuals of any effective remedy, in breach of Art. 47 of the Charter of Fundamental Rights of the EU.

As regards duration, in cases not involving a threat to public order or security, the Commission extends the maximum length of an entry ban from five to ten years, while the Council further increases it to twenty years (Art. 10(6)). Such durations may severely interfere with the right to private and family life under Art. 8 ECHR, particularly for individuals who have established strong personal and social ties in the EU prior to removal. Although the Council retains the possibility to withdraw, suspend or shorten an entry ban in justified individual cases (Art. 11(2)), it significantly restricts individuals’ ability to request such measures (Art. 11(3)).

8) Persons posing “security risks”: the Commission introduces a distinct category of persons subject to return to whom heightened coercive measures may apply: those deemed to pose “security risks” (Art. 16). The Council further expands this already vaguely defined category. In particular, it includes within the notion of a threat to public policy or security and national security persons subject to an alert in the Schengen Information System (SIS), which itself encompasses situations as minor as attempts to circumvent rules on entry or residence. The Council also defers to national law for defining “other security threats,” further diluting legal certainty. Persons falling within this category may be subjected to additional coercive measures, including detention—which may be carried out in prisons and potentially be of indefinite duration—and entry bans exceeding twenty years. This approach blurs the distinction between criminal and migration law, reinforces harmful stereotypes linking migration and security risks, and circumvents the procedural safeguards that apply in criminal proceedings. While States have a legitimate interest in protecting public order and national security, criminal law should apply equally to all persons under a State’s jurisdiction. Using migration law as a substitute for criminal justice undermines fundamental legal principles and the rule of law.

9) “Return hubs”: among the Commission’s so-called “innovative solutions” to increase return rate is the establishment of “return hubs.” The proposal allows Member States to remove individuals to a third country with which they have concluded an agreement or arrangement (Art. 17). The person need not have any link to that country, nor must the country qualify as a safe third country, although certain conditions are envisaged. The Council’s position clarifies that such agreements or arrangements may regulate the modalities of onward return and that detention may result from their implementation (Art. 17(2a)). It also relaxes notification obligations vis-à-vis the Commission and other Member States (Art. 17(3)), reducing transparency and oversight.

While the externalisation of returns may appear attractive to Member States, recent migration arrangements show that such models often entail significant legal challenges, operational difficulties and high financial costs for the outsourcing State. More fundamentally, return hubs raise serious human rights concerns, in particular regarding detention and onward return. States cannot evade their international obligations by outsourcing migration control. Depending on the degree of involvement, responsibility for violations may be attributed to the externalising State, the third State, or both, including through complicity in internationally wrongful acts.

10) Children: the Council reduces child-specific safeguards throughout the proposal. It weakens protections relating to age assessment (Art. 19), the appointment and role of a guardian for unaccompanied children (Art. 20(2)), and legal assistance and representation (Art. 25). The Council also removes the provision in the current Directive (Art. 10(1)) that requires assistance for an unaccompanied child by bodies other than the authorities enforcing return to be granted before issuing a return decision. This safeguard is crucial to ensure that the best interests of the child are properly assessed prior to any return decision. Under international law, children may be returned only where this is in their best interests. The Council further removes schooling from the explicit grounds for extending the period for voluntary departure (Art. 7(1c)). It also deletes the Commission’s proposed exemption of families with children from return to “return hubs” (Art. 17(4)), leaving only unaccompanied children excluded from this measure. Moreover, children are not exempt from newly introduced coercive measures, including cooperation obligations and sanctions for non-compliance (see above (5)), as well as investigative measures (see above (6)).

As regards detention, both the Commission and the Council disregard international standards according to which children should not be placed in immigration detention at all. The expanded detention regime (see above (4)) applies equally to children. The Council further removes the mandatory ex officio review of detention of unaccompanied children (Art. 33(2)) and relaxes requirements concerning conditions and services in detention, including access to education (Art. 35(2)–(3)). These changes are incompatible with the best interests of the child and with children’s rights to liberty and education under the UN Convention on the Rights of the Child.

 

Concluding thoughts

The Council: as this assessment demonstrates, the Council’s position lacks proportionality and opens the door to arbitrary measures. Human rights safeguards are systematically weakened. Even the fundamental, non-derogable prohibition of refoulement no longer prevents the issuance of a return decision. The accumulation of multiple—and at times overlapping—coercive measures, including far-reaching cooperation obligations and sanctions for non-compliance, runs counter to the stated objective of the reform, namely to streamline and simplify return procedures. Moreover, the Council’s frequent deferrals to national law—including on core issues such as grounds for detention, criteria for establishing a risk of absconding, and the definition of a security threat—undermine the very rationale for replacing a directive with a regulation and defeats the goal of harmonisation of return procedures.

Other players: the Council’s position was made possible by the Commission. The Commission has consistently relied on the narrative of a low return rate, while maintaining a taboo on non-return and regularisation, despite having funded multiple research projects that provided evidence on this matter. The Commission’s proposal itself already contained several measures raising serious concerns from a human rights perspective. By advancing this proposal in a context of political polarisation on migration—where centrist parties increasingly adopt right-wing and enforcement-driven approaches—the Commission has fallen short of its role as guardian of the Treaties and defender of the Union’s fundamental values. Attention now turns to the European Parliament, with the hope that its final report will offer a human-rights-compliant position to restore a measure of balance before inter-institutional negotiations begin.

First they came for migrants: the Council’s position reflects the view, advanced by some States, that the European Court of Human Rights—or human rights more broadly—has become excessively protective of migrants in an irregular situation and unduly restrictive of States’ prerogatives to expel them. To be clear, any attempt to narrow the scope or application of human rights conventions in relation to migrants constitutes an attack not only on migrants’ rights, but on the rights of all persons. Once human rights are weakened for one group deemed politically expendable, derogations and qualifications can be expected to follow for others. What is presented as an exceptional response to migration thus risks becoming a broader erosion of the universality and indivisibility of human rights protection.

 

 

From Large Language Models to Long Lasting Manipulations: the AI Act and generative AI advertising

 

Annelieke Mooij*  and Anuj Puri**

*Assistant Professor at the Public Law & Governance Department, Tilburg Law School

** Post-Doctoral Researcher at the Public Law & Governance Department, Tilburg Law School

Photo credit: piqsels, via Wikimedia Commons

 

Introduction

General purpose Large Language Models (LLMs) are amongst the most discussed innovation of the century, with AI developers even being named persons of the year by the Times magazine. Amongst the leading General purpose LLMs, perhaps the most famous one is ChatGPT – which is offered by Open AI. In light of such success, it may be surprising for many to learn that Open AI operates at huge losses. Its annual revenue is predicted at 13 billion dollars, which suffices to only a fraction of its computing costs which totals approximately 1.4 trillion dollars over the next eight years. It was therefore not entirely unexpected that Open AI was preparing ChatGPT for the inclusion of advertisement. The potential introduction of such advertisements raises significant ethical and legal concerns.

Consider the following excerpt from ChatGPT’s Memory FAQ,

ChatGPT can remember useful details between chats, making its responses more personalized and relevant. As you chat with ChatGPT, whether you’re typing, talking, or asking it to generate an image, it can remember helpful context from earlier conversations, such as your preferences and interests, and use that to tailor its responses.

Depending upon one’s penchant towards customization or preference for privacy, such features may either increase usability or raise privacy concerns or both. The inclusion of advertisement within General purpose LLMs should, however, concern even the least privacy conscious users.

ChatGPT has already taken steps to provide a user with customized in-conversation and instant check-out shopping, thereby creating new potential avenues for manipulation of consumers. Hence, it is not surprising that its plan to introduce ads was met with backlash. Most of the critique, however, seemed focused on the inclusion of advertisement in ChatGPT pro-plans and the lack of quality of the suggested ads. The advertisement plans have purportedly been currently put on hold to improve ChatGPT’s core features including personalization. It is not unlikely that ChatGPT may roll out an improved version that includes personalized ads. Hence, there exists an urgent need to examine the possibility of such advertisements manipulating consumers.

Manipulation Risks

Consider a potential scenario where an individual is in distress over the fall out of a personal relationship and reaches out to a general purpose LLM like the ChatGPT for advice. The LLM responds by advising the user to spend time and money on self-care by shopping for products such as clothes, shoes etc. with “helpful” links to shopping websites and perhaps a “helpful” image of the product. The user’s prior usage history may lead to their vulnerable situation being exploited for surveillance capitalist purposes. Such plausible uses of user’s information by the firms developing and deploying general purpose LLMs raise concerns pertaining to the use of manipulative techniques.

              From an ethical perspective, manipulation can be understood in various ways— such as manipulation in the form of introduction of non-rational influence (which renders it closer to subliminal technique), manipulation as a form of pressure, and manipulation as a form of trickery (which is conceptually linked to deception). Susser et al have defined manipulation as imposing a hidden or covert influence on another person’s decision-making and offered a widely accepted account of online manipulation as the use of information technology to covertly influence another person’s decision making. Manipulation understood in this manner raises concerns pertaining to the covert exploitation of a LLM user’s emotional vulnerabilities for commercial exploitation purposes. Before we address the question of existing legal remedies, it would be helpful to highlight some of the backdrop conditions which pave way for the potential manipulation of the consumers.

              Two common conceptual concerns lie in the backdrop of the purported use of manipulative AI— trust and anthropomorphization. The propensity of users to trust general purpose LLMs with queries pertaining to all aspects of their lives, even when they are not trustworthy, is at the heart of the  manipulation risks. Secondly, the conversational nature of the interaction with the LLM increases the odds of the user getting exploited on account of the tendency to anthropomorphize such interactions. It is worth noting that the undeserved inducement of trust and anthropomorphization are borne out of the design choices made by the developers. The potential to rely on previous conversations, the covert nature of the exercised influence, trusting propensity of the users along with the tendency to anthropomorphize the conversation lead to a fertile ground for potentially long-lasting manipulation of the user. This is where the legal remedies provided under the EU AI Act have an important role to play in protecting vulnerable users.

Legal Remedies

Article 5 of the AI-Act prohibits the deployment of manipulative AI. It is, however, difficult to define what constitutes manipulation. According to the Commission’s Guidelines on the AI Act, “[m]anipulative techniques are typically designed to exploit cognitive biases, psychological vulnerabilities, or situational factors that make individuals more susceptible to influence” This raises the question when ChatGPT’s advertisement exploits a psychological vulnerability and/or situational factor. And whether legal distinctions of vulnerabilities can reasonably be made.

One way of addressing the question of vulnerability is by examining the tendency to anthropomorphize general purpose LLMs. Users trust such LLMs as confidants instead of realizing that their data is being used for commercial exploitation. In view of such tendencies and dependencies, one could argue that general purpose LLM advertisements are inherently exploiting the vulnerability of the users. Thus, such advertisements are manipulative by design. This, however, fails to recognize that some users may only use the LLM as a search engine. This ambiguity in usage demonstrates the legal conundrum surrounding the identification of vulnerability.   

When it comes to consumer protection, the question of exploitation of vulnerability has been addressed in the Unfair Commercial Practices Directive. In consumer cases, the Court of Justice of the EU has held that the in order to be considered unlawful advertisement should manipulate a reasonably informed and circumspect consumer. The average consumer is an interpretative standard that the CJEU develops based on the product as an expression of proportionality. The average consumer is defined in relation to a product’s target audience, certain groups, such as children, are considered inherently more vulnerable. Gaming platforms for children, for instance, must therefore comply with stricter advertisement rules. From the perspective of vulnerability determination on the basis of product, it is an open-ended question what does being a reasonably informed and circumspect user of AI entail? Should it be assumed that AI users always have a minimum level of knowledge that all their interactions with the LLMs are aimed at commercial gain? Should the reasonable consumer be circumspect that all prompts are potential data fodder for exploiting (future) vulnerabilities; and be suspicious of all AI results at all times? Even when they look up the recipe for apple pie? There are some who would argue that advertisement based on algorithms and big data are inherently manipulative. If we accept this argument, general purpose LLMs should not be able to include any form of advertisement.

As stated before, development and deployment of AI systems is currently extremely resource intensive. A proponent of inclusion of advertisement in general purpose LLM may therefore argue that ad-driven revenue generation model reduces digital exclusion. This argument however begs the question whether access to AI systems in garb of exploitation of user’s vulnerabilities through manipulation is equitable access at all. A more sustainable solution is perhaps not to prohibit advertisement, but to regulate against exploitation. This, however, requires a shift in approach.

A possible solution could be to train AI systems to differentiate between (extremely) vulnerable prompts (questions) such as how to deal with a break-up and prompts with lower vulnerability such as how to bake an apple pie. This would require a shift in perspective. Rather than defining the average consumer, it would require defining the average prompt or AI interaction, whereby prompts such as “how to get over a break-up” indicate a vulnerability that is legally protected from exploitation. However, such a classification attributes the power to AI systems to distinguish between users that are in a potentially vulnerable  state and those that are not. Even if such a hypothetical position were to be possible, it would not address all the underlying ethical and legal concerns. Algorithmic determination of vulnerability is as likely to reflect the normative choices made by the developers and the computational trade-offs made in the training data sets. It is unlikely that these accurately reflect vulnerabilities without bias, as development of AI systems is not known to reflect diversity.

Another avenue to explore is not to regulate the prompts, but the amount of personal history that general purpose LLMs may access to generate advertisement. Such regulation would do justice to the argument that big data & algorithm is inherently manipulative. Limiting the amount of data that can be used for advertising has the additional advantage of clarity. If for instance only ten data points can be used for advertising, this provides legal certainty. The difficulty however is enforcement – as it requires verifying source codes. Further it is difficult to construe a safe harbour provision for data collection, depending upon the nature of data points, even limited data can be used for undermining an individual’s autonomy. Furthermore, it does not reflect the reality of people who use a LLM as a confidant, even though it is not trustworthy, on account of the seemingly anonymized and private interaction with the AI system. Thus making them vulnerable towards its manipulative influences.

Questions for the future remain

While the potential introduction of advertisements in general purpose LLMs such as the ChatGPT might have been paused for now, the financial incentives to justify the Silicon Valley optimism remain, which are currently also driving the policy measures across the Atlantic. The advent of these potential advertisements would not be the last attempt to test the regulatory resolve in the technological battle to impinge upon human autonomy. But by taking a strong stand, and withstanding the geo-political pressure, the EU institutions can make it amongst the first red lines that should not be crossed.

 

 

Asylum Pact 2.0: the EU moves towards more stringent rules on ‘safe third countries’ and ‘safe countries of origin’

 

Steve Peers, Professor of Law, Royal Holloway University of London

Photo credit: Steve F, via Wikimedia Commons

 

Introduction

The EU’s asylum pact, agreed in 2024, is mostly not fully in force yet; it will largely apply from 12 June 2026. But even before that date, the EU is planning to make its rules more stringent – and even to apply some of them in advance. The prospect of these planned changes moved closer recently, as both the European Parliament and the EU Council, ie Member States’ ministers firmed up their negotiation positions on two separate proposals, and may negotiate an agreed final text of them both in the near future.

This blog post summarises the proposals in their context, and then examines the proposed amendments coming from the EP and the Council, concluding with an assessment of the main issues arising from the negotiations – including the prospect of a ‘Rwanda clause’ in EU law, mirroring the last UK government’s attempt to designate that country as ‘safe’ even for asylum-seekers who had not travelled through it, and the European Parliament’s suggestion to curtail judicial review in a way that would obviously breach the EU’s Charter of Fundamental Rights.

 

Background

The EU asylum pact includes a Regulation on asylum procedures (which I previously discussed here), replacing the current Directive on asylum procedures, dating from 2013, which sets out detailed rules on the processing of asylum applications. These include rules on both supposedly ‘safe countries of origin’ (countries which asylum-seekers originate from) and supposedly ‘safe third countries’ (countries other than the countries of origin, which asylum-seekers ‘should’ apply for asylum in instead).

Variations of these two concepts have existed in EU law for a long time, dating back initially to 1992 in the form of the London Resolutions of Member States’ ministers (see here and here). The concepts then appeared in the initial 2005 asylum procedures Directive, since replaced by the 2013 version of the rules. Although the asylum pact Regulation adopted in 2024 made these rules more restrictive for asylum-seekers, this was deemed insufficient, hence the move to change the rules in the pact already.

 

‘Safe countries of origin’

Current rules

The 2013 rules provide that a country can be considered a ‘safe country of origin’ for non-EU asylum seekers if it meets specified human rights standards, taking account of country of origin information from various sources such as the UNHCR and EU asylum agency. Use of this rule is currently an option for Member States. The rule can only apply if an asylum-seeker is a national of the country concerned, or is a stateless person formerly habitually resident there. Member States must lay down further rules in national law if they use the concept. They must also regularly review the list of countries concerned and inform the Commission of the list.

Unlike the 2005 rules, it is not possible to define a country as safe only in part, either geographically or for selected groups of people, as the Court of Justice has recently confirmed (see here and here). These judgments put a spanner in the works – at least temporarily – of the Italy/Albania arrangements, under which Italy planned to remove to Albania select groups of asylum-seekers who came from a ‘safe country of origin’, which was designated as ‘safe’ for only some groups of people; the asylum-seekers were to remain in Albania while Italy processed their application. In the latter judgment (Alace), the Court of Justice also said that: a Member State had a choice of routes to designate a ‘safe country of origin’ in national law, but its designation must be subject to judicial review; the sources of information used for the designation must be provided to the applicant and to courts; and courts must be able to examine other sources of information.

Defining a non-EU country of origin as ‘safe’ creates a presumption, which can be rebutted by the asylum-seeker if they can show ‘any serious grounds for considering the country not to be a safe country of origin in his or her particular circumstances and in terms of his or her qualification as a beneficiary of international protection’. It is also a ground for fast-tracking the consideration of asylum applications, although the current rule does not specify a time limit to this end (just that it should be ‘reasonable’, and can be exceeded if necessary to examine the application properly). In comparison, the time limit for ‘standard’ asylum applications in the current law is six months (although there are several grounds for extending that deadline).

The ‘safe country of origin’ rule is also one ground for applying a fast-track border procedure to asylum-seekers’ applications; again, such procedures are currently optional. Here there is a specific deadline, namely to conclude the process within four weeks, otherwise the asylum-seeker must be allowed on to the territory, and (as confirmed by Court of Justice case law) released from detention if that was the only ground for it. Finally, Member States may deny suspensive effect appeals in ‘safe country of origin’ cases, although asylum-seekers must have an opportunity to request a court to grant such suspensive effect.

2024 Regulation

Under the 2024 asylum pact Regulation, applicable (as things now stand) to asylum applications made after 12 June 2026, the ‘safe country of origin’ rules become mandatory for Member States. (Most, but not all, Member States were applying them already) The basic definitions of human rights standards in the country of origin remain, as do the rules on the sources of data to be considered. But under the new rules, it will be possible to designate a non-EU country of origin as ‘safe’ in part, either geographically (exceptions for ‘specific parts of its territory’) or for some groups of people (‘clearly identifiable categories’). Conversely, there is nothing to suggest that the other aspects of the Alace judgment (on effective judicial review of designations of ‘safe third countries’) cease to apply.

One new aspect of the rules is that it is possible for the EU to adopt its own common list of ‘safe countries of origin’, via the ordinary legislative procedure. This is accompanied by rules on dropping countries from the list in the event of ‘significant changes’ there, initially by means of a delegated act adopted by the Commission, then via legislation. Member States cannot put the country back on their national list of ‘safe countries of origin’ while a delegated act suspending it from the list applies; for two years after the country is dropped from the common EU list via legislation, Member States need the Commission’s approval to reinstate it on their national lists.

A similar new ground for accelerated procedures has been added: Member States must also fast-track cases where the most recent annual recognition rate for asylum claims for the asylum-seekers’ nationality is below 20% EU-wide at first instance (ie before appeals), ‘unless the determining authority assesses that a significant change has occurred in the third country concerned since the publication of the relevant Eurostat data or that the applicant belongs to a category of persons for whom the proportion of 20 % or lower cannot be considered to be representative for their protection needs, taking into account, inter alia, the significant differences between first instance and final decisions.’

There will now be a specific three-month deadline to decide on accelerated cases, although an authority can decide to consider the asylum application on the merits if the case is too complex. (The general deadline to decide on applications remains six months; the possibilities of extending that deadline have been partly curtailed).

‘Safe country of origin’ remains a ground (alongside now the ‘20% recognition rate’ rule) for considering applications in a border procedure, but that procedure has been overhauled: it is now mandatory for Member States for a certain number of asylum-seekers, provides for a longer period of application (12 weeks, now including appeals), and is subject to more exceptions. In particular, unaccompanied minors can no longer be subjected to it, except where they are ‘national security’ or ‘public order’ risks.

As for appeals in ‘safe country of origin’ cases, it is now the standard rule that they do not have suspensive effect (except for unaccompanied minors in the context of the border procedure), although as before it must be possible for asylum seekers to request suspensive effect from the courts. Some additional safeguards that currently apply to the lack of suspensive effect in border procedure cases have been dropped.

Commission proposal

The Commission’s proposal, dating from April 2025 (see my previous comments here), would first of all allow (as an option) the early application of the revised rules on ‘safe country of origin’, as well as the new ‘20% recognition rate’ rule, in order to provide for the earlier application of the Italy/Albania arrangements; other Member States might have a use for earlier application of the rules too. In fact it would also allow for early application of the partial designation of countries as ‘safe third countries’ too. It will also allow application of the 2024 version of these special rules in the current version of the border procedure (among other things, the 2024 exclusion of most unaccompanied minors from the border procedure will not apply).

Secondly, it would set out a common EU list of ‘safe countries of origin’, as from the entry into force of the Pact: seven named countries (Bangladesh, Colombia, Egypt, India, Kosovo, Morocco and Tunisia) plus candidates for EU accession (Serbia, Montenegro, Albania, North Macedonia, Bosnia-Herzegovina, Turkey, Ukraine, Moldova and Georgia). The accession candidates would be subject to special rules: their listing would switch off if they have been subjected to EU foreign policy sanctions ‘adopted in view of the country’s actions’, or if their recognition rate is above 20% at first instance, or if there is a ‘serious and individual threat to a civilian’s life or person by reason of indiscriminate violence’ in that country (which is one of the grounds for ‘subsidiary protection’ in EU law). It is not certain whether the usual rules on suspending a country from the list also apply to accession candidates, given that a special new category has been created for them. The proposal would list all these countries as a whole, ie not using any of the exceptions for parts of a country or groups of people which the Commission argued were so essential to provide for Member States.

Council position

The recently agreed Council position has taken over most of the Commission proposal. While the Council accepts the early application of parts of the Regulation and the common EU list of ‘safe countries of origin’, including the accession candidate countries (subject to a special rule), it would make a few amendments. (A statement by Hungary objects to having a special rule for accession candidates, preferring to list them automatically without any conditions – although listing them automatically is a special rule in itself)

First of all, the Council position would amend the proposal so that it would be possible to suspend a country partly from a common EU list of ‘safe third country’ or ‘safe country of origin’, on the same basis that Member States can designate a country only partly, ie geographically or as regards groups of people. If a country is partly removed from the common EU list by EU legislation, Member States would not need the Commission’s approval to reinstate that country partly to a national list.  

Secondly, the Council would alter the special conditions applicable to accession countries. The ‘subsidiary protection’ ground for disapplying the status would no longer refer to an ‘individual’ threat, presumably because an assessment is being made of a country as a whole in this context. Also, the foreign policy sanctions ground for disapplying the status would only apply to that country’s actions ‘affecting fundamental rights and freedoms that are relevant for the criteria of designation of a third country as a safe country of origin’ set out in the Regulation. The Council also sets out a procedure for applying these exceptions: the Commission must inform Member States and the Council of the change in status. However, as regards the (quasi-)‘subsidiary protection’ exception, the Commission needs the prior approval of the Council (presumably by qualified majority) before informing Member States of that change. According to the preamble, this is necessary in light ‘of the potential implications for the external relations of the Union and the Member States’ in this scenario.

European Parliament position

The European Parliament’s position (agreed by a committee, and to be reviewed in the full Parliament shortly) is similar to the Council’s. On the first point, the EP agrees that it would, in effect, be possible for countries on the common EU list to be suspended only partly, although it goes into less detail than the Council does.

On the second point, the EP retains the ‘individual’ threat aspect of removing an accession candidate from the common EU ‘safe third country’ list, but also adds that a candidate country should be automatically removed from the list on this ground if the EU’s temporary protection Directive has been applied to that country. This obviously refers solely to Ukraine at present (until March 2027, as things stand). The EP would amend the foreign policy sanctions ground for removal from the list in the same way as the Council.

As for the process, the EP wants the Commission to remove candidate countries from the list by means of a delegated act, rather than by informing the Council and getting the Council’s approval in some cases – although the Council has a role anyway in scrutiny of delegated acts (and in invoking and extending the application of the temporary protection Directive). Using a delegated act – which is, after all, already the usual process set out in the Regulation for suspending a country from the common list – would also give the EP a role in the suspension process.

The EP would also make some amendments to the preamble. One of them, indicating that assessments of the safety of non-EU countries should be ‘accessible’, reflects the Alace judgment. But another amendment to the preamble plainly conflicts with that judgment, purporting that:

…national judicial review should examine the detailed evidence regarding an applicant’s individual situation justifying, in his or her case, the inapplicability of the concept of safe country of origin and not the designation as such.

 

‘Safe third countries’

Current rules

The 2013 rules (again, an option for Member States) provide that a non-EU country can be considered a ‘safe third country’ (the position of EU countries and the associated countries of Switzerland, Norway, Iceland and Liechtenstein is a separate issue) if several criteria are satisfied: no threat to [life or liberty’ on any of the grounds set out in the Refugee Convention; ‘no risk of serious harm’ as defined in EU law as regards subsidiary protection; respect of the non-refoulement principle in the Refugee Convention (ie not sending the asylum seeker to an unsafe country); respect for ‘the prohibition of removal, in violation of the right to freedom from torture and cruel, inhuman or degrading treatment as laid down in international law’; and the possibility ‘to request refugee status and, if found to be a refugee, to receive protection in accordance with’ the Refugee Convention.  

The ‘safe third country’ principle must be ‘subject to rules laid down in national law, including’: rules requiring a ‘connection’ with the other country, ‘on the basis of which it would be reasonable’ for the asylum-seeker to go there; rules on ‘methodology’ concerning the application of the principle to particular countries or applicants; and rules permitting the asylum-seeker to challenge the alleged safety of the country concerned for them, as well as their supposed connection with it.

‘Safe third country’ applications may be considered inadmissible, ie not considered on the merits; but if the supposedly safe country does not permit the person concerned to enter its territory, the Member State must fully consider the merits of their claim. This rule has been confirmed by the Court of Justice, in a case where Greece was breaching it by ruling thousands of claims inadmissible because Turkey was ‘safe’, even though Turkey was no longer readmitting any of the asylum seekers concerned.

A special inadmissibility interview is held, rather than an interview focussed on the well-foundedness of the asylum claim as such. The Directive hints that Member States may have separate time limits for such cases.

As with ‘safe countries of origin’, the ‘safe third country’ rule is another ground to apply a special borders procedure (if Member States opt to do so). Conversely, the suspensive effect of an appeal cannot be denied in ‘safe third country’ cases.

2024 Regulation

The 2024 rules – which remain an option for Member States – are now subject to a definition of ‘effective protection’, in place of an opportunity to request and receive Refugee Convention refugee status: if a non-EU country ‘has ratified and respects’ the Convention, within the limits of any reservations and limitations, that country ‘shall be considered to ensure effective protection’; but if has not ratified the Convention, or applies a geographical limit to it (ie Turkey), that country ‘only’ offers effective protection for people where, ‘as a minimum’, that country allows people to remain, offers subsistence, health care and education, and ‘effective protection remains available until a durable solution can be found’. So even countries which have not ratified the Convention at all can be regarded as offering ‘effective protection’, if they meet these other conditions.

As with ‘safe countries of origin’, it is now possible to regard a country as a ‘safe third country’ only partly, ie ‘with exceptions for specific parts of its territory or clearly identifiable categories of persons’. A country can even be a ‘safe third country’ for an individual applicant. There is a special safeguard for unaccompanied minors:

A third country may only be considered to be a safe third country for an unaccompanied minor where it is not contrary to his or her best interests and where the authorities of Member States have first received from the authorities of the third country in question the assurance that the unaccompanied minor will be taken in charge by those authorities and that he or she will immediately have access to effective protection as defined in [the Regulation].

The law still rules out holding ‘safe third country’ applications to be inadmissible if the other country refuses to admit or readmit the person concerned on to its territory.

There will now be a two-month deadline for deciding ‘safe third country’ applications (and most other inadmissible applications), with a possible extension of up to two months in certain circumstances.

The revised border procedure in the Regulation (described above), also applies to ‘safe third country’ cases. There is also now a new prospect of a common EU list, which works the same way as the possible common list of ‘safe countries of origin’ (and removals from it) described above. Finally, appeals in ‘safe third country’ cases still have automatic suspensive effect.

Commission proposal

The Commission’s proposal, dating from May 2025 (see my previous comments here), would alter the rules as regards the asylum-seeker’s links to the country concerned, so that a ‘connection’ to that country would no longer be the only ground for applying the principle. It could also apply where either the asylum seeker had transited through that country, or a Rwanda-type deal existed with a country that the asylum had neither a connection with nor transited through: ‘there is an agreement or an arrangement with the third country concerned requiring the examination of the merits of the requests for effective protection made by applicants subject to that agreement or arrangement’. However, the latter new criterion would not apply to unaccompanied minors, and in any case there would be an obligation to consider the best interest of the child. Member States would have to inform the Commission and other Member States in advance of concluding such agreements or arrangements.

Secondly, the Commission proposal would alter the rules on appeals, so that there would no longer be automatic suspensive effect in ‘safe third country’ appeals, except for unaccompanied minors subject to the border procedure. Nevertheless, it would be possible to request a court to grant suspensive effect.

The ‘safe third country’ principle would remain optional for Member States, and the Commission does not propose to change the criteria defining the ‘safety’ of a country as such, the applicable deadlines, the related rules on the border procedure, the current safeguard for unaccompanied minors or the requirement that Member States must consider the merits if the third country concerned refuses to admit the asylum seeker. Nor does it propose to use the power to adopt a common EU list of ‘safe third countries’.

Council position

The Council’s position would take on board the Commission’s main points, clarifying that the transit in question must be ‘on the way to the Union’, and providing for the EU, not only Member States, to enter into negotiations with non-EU countries which the asylum seekers have no link at all to. New provisions would require the Commission, when negotiating such agreements or arrangements, to take Member States’ existing agreements or arrangements into account, including the potential impact of EU agreements or arrangements on cooperation of the non-EU countries with certain Member States. But an agreement or arrangements with the EU, once concluded, will take priority over agreements or arrangements with Member States, where they are incompatible.

Member States negotiating such agreements or arrangements must inform other Member States and the Commission of them prior to their entry into force or provisional application, and also inform other Member States and the Commission of any amendments or termination. A notification should come at an (unspecified) earlier point where the agreements or arrangements are with a non-EU country that borders on a Member State (ie Germany sending asylum seekers to Turkey may have a particular impact on Greece). As an option (according to the preamble), Member States could consult the Commission at an earlier stage, with a view to considering the compatibility of the draft agreement or arrangement with EU law. The Council version would also allow for different ways to inform a non-EU country that the applications of asylum-seekers being sent there have not been considered on the merits.

This negotiation position maintains the exclusion of unaccompanied minors from the ‘Rwanda clause’. On the other hand, it would drop the general reference to the rights of the child that the Commission wanted to insert in the main text, although the preamble would retain a reference to this principle, adding that ‘Member States should also take due account of the principle of family unity when applying the safe third country concept’. The safeguard already in the 2024 Regulation for unaccompanied minors, described above, would not be amended by either the Commission or the Council.

The Council’s version would also drop automatic suspensive effect in cases where the asylum seeker has international protection from another Member State. Greece objects to this. It should be noted that the Court of Justice case law provides for the possibility of applying for international protection in another Member State to avoid harsh conditions in the Member State which granted such protection, where the latter Member State treats the beneficiaries of international protection so badly that it amounts to a breach of the EU Charter of Fundamental Rights.

Finally, the Council version tries to clarify some elements of the ‘safe third country’ rule in the preamble. As regards the concept of a ‘connection’ (which will, of course, matter less than it does at present, given the two new categories of ‘safe third countries’):

While taking fully into consideration the parameters outlined in the case law of the Court of Justice of the European Union, Member States should be able to apply the safe third country concept on the basis of a connection as defined in conformity with national law or practice, in so far as specifically defined therein. The connection between the applicant and the third country could be considered established in particular where members of the applicant’s family are present in that country, where the applicant has settled or stayed in that country, or where the applicant has linguistic, cultural or other similar ties with that country.

Transit through a non-EU country is also clarified:

transit through a third country could include the situation where an applicant has passed through, or stayed on, the territory of a third country on the way to the Union, or where the applicant has been at the border or in a transit zone of a third country, where he or she has had the possibility to request effective protection with the authorities of that country

As for the Rwanda-style arrangements, they ‘could include a variety of case-processing modalities, such as simplified, group or prima facie procedures’. This wording seems to imply a potentially dismissive attitude to the non-EU country considering the merits of applications.

European Parliament position

The European Parliament’s position (again agreed by a committee, and to be reviewed in the full Parliament shortly) is similar to the Council’s. In fact, the EP would only make two changes to the main text of the Commission proposal. First, like the Council, the EP would also provide for the possibility of the EU, not only individual Member States, negotiating Rwanda arrangements, although it goes into less detail than the Council version about the mechanics of this. (Its proposed changes to the preamble would, however, require Member States to inform about bilateral talks at an earlier point, and inform the EP too; and the arrangements would have to be in writing).

Second, the EP version would subject unaccompanied minors to the Rwanda clause if there are ‘reasonable grounds’ to believe that they are a security or public order threat ‘under national law’. While the Court of Justice has usually interpreted security exceptions from asylum law narrowly, the reference to national law may be intended to give Member States more leeway. The EP would, however, retain the general reference to the ‘best interests of the child’ being added to the main text.

The EP also suggests changes to the preamble to the proposal, some of which align broadly with the Council’s. On the ‘connection’ with the ‘safe third country’, the EP states that:

The connection between the applicant and the safe third country could be considered established in particular where members of the applicant’s family are present in that country or where the applicant has settled or stayed in that country, or where the applicant has other links with that country, such as the same or similar language, or other economic, cultural, religious, or geographical links.

Unlike the Council’s version, there are, however, no broad references to national law in this context.

The EP version supports automatic suspensive effect being removed from most ‘safe third country’ appeals, although unlike the Council, it would not also remove automatic suspensive effect from appeals made by those who have international protection already from another Member State.

 

Assessment

First and foremost, one of the amendments proposed by the EP would be plainly unlawful, as an obvious breach of the EU Charter of Fundamental Rights. The Court of Justice’s Alace judgment refers several times to Article 47 of the Charter (the right to an effective remedy and a fair trial), when ruling that it must be possible to challenge the designation of countries as ‘safe countries of origin’ (underlining and bold text added):

…it should be noted that the obligation imposed on Member States by Article 46(1) of Directive 2013/32 to provide for a right to an effective judicial remedy for applicants for international protection, the scope of which is defined in Article 46(3) of that directive, corresponds to the right to an effective remedy guaranteed by Article 47 of the Charter… [para 53]

…although, in the absence of EU rules on the matter, it is for the national legal order, in accordance with the principle of procedural autonomy of Member States and subject to the observance of the principles of equivalence and effectiveness, to lay down the detailed procedural rules governing remedies for ensuring that individual rights derived from the EU legal order are safeguarded, Member States nevertheless have the responsibility to ensure observance in every case of the right to effective judicial protection of those rights as guaranteed by Article 47 of the Charter, the scope of that right being clarified, in the present case, by Article 46 of Directive 2013/32… [para 64]

…the choice, by a Member State, of the competent authority and the legal instrument effecting the designation, at national level, of safe countries of origin, in accordance with Articles 36 and 37 of Directive 2013/32, cannot affect its obligations under that directive. It is thus for each Member State, inter alia, to ensure respect for the right to an effective judicial remedy which Article 46(1) of that directive confers on applicants for international protection against decisions taken on their applications, the scope of which is defined by Article 46(3) of that directive. [para 65]

In that regard, the Court has held that, in accordance with Article 46(3) of Directive 2013/32, read in the light of Article 47 of the Charter, where an action is brought before a national court or tribunal against a decision taken on an application for international protection – examined in the context of the special scheme applicable to applications lodged by applicants from third countries designated, in accordance with Article 37 of that directive, as safe countries of origin – that court or tribunal must, as part of the full and ex nunc examination required by Article 46(3) of that directive, raise, on the basis of the information in the file and the information brought to its attention during the proceedings before it, a failure to have regard to the material conditions for such designation, set out in Annex I to that directive… [para 66]

Consequently, and having regard to the case-law referred to in paragraphs 62 and 63 above, the fact that a Member State has chosen to designate safe countries of origin by means of a legislative act cannot be such as to preclude the national court or tribunal seised in the circumstances set out in the preceding paragraph of the present judgment from reviewing, even if only indirectly, whether the designation of the third country in question as a safe country of origin complies with the material conditions for such a designation, set out in Annex I to Directive 2013/32. [para 67]

In the light of the foregoing, the answer to the first questions is that Articles 36 and 37 and Article 46(3) of Directive 2013/32, read in the light of Article 47 of the Charter, must be interpreted as not precluding a Member State from designating third countries as safe countries of origin by means of a legislative act, provided that that designation can be subject to judicial review as regards compliance with the material conditions for such a designation, set out in Annex I to that directive, by any national court or tribunal hearing an action brought against a decision taken on an application for international protection, which had been examined under the special scheme applicable to applications lodged by applicants who are from third countries designated as safe countries of origin. [para 68]

As the Charter has the same legal value as the Treaties (Article 6 TEU), any EU law adopted in breach of it would be invalid.

The EP majority has also not thought this amendment through. It is not reflected in the main text of the Regulation; and it is unclear if the amendment is somehow intended to prevent a review of the validity of a designation on the common EU list too. But Article 267 TFEU provides that a national court can ask the Court of Justice about the validity of EU legislation; if the EP amendment is intended as an attempt to preclude that, then it would be unlawful for a second reason.

Otherwise, as discussed already, there is not must difference between the EP and Council positions on either proposal. Traditionally the EP has taken a significantly more liberal view than the Council on asylum issues, but now the two are broadly in sync (and both in a more restrictive direction than in the past), with the EP even more restrictive than the Council on some points, although the Council is more restrictive than the EP on some points too.

On the ‘safe country of origin’ proposal, the EP’s position on the definitions and process regarding candidate countries is more convincing: it is logical that applying the temporary protection Directive should lead to an automatic exclusion from the common list of ‘safe countries of origin’, and it would make sense to follow the usual delegated acts process for suspending a country from the list, rather than an ad hoc intergovernmental process that only gives a role to the Council (there’s history here: the Court of Justice previously ruled against the Council’s botched attempt at an ad hoc intergovernmental process as regards the very same issue).

As regards the ‘safe third country’ proposal, the Council’s attempt to extend the removal of automatic suspensive effect is an unprincipled reach into another area of EU asylum law, and would in any event remove an essential feature of an effective remedy as regards potential breaches of Charter rights. But the removal of automatic suspensive effective from ‘safe third country’ cases is also problematic, especially in light of the extremely broad definition of the concept that would follow from the proposals.

The introduction of a ‘Rwanda clause’ in EU asylum law undercuts the traditional argument that asylum-seekers ‘should have’ applied in another country. Nevertheless, this rationale even appears in the Council’s press release:

The safe third country concept allows EU member states to reject an asylum application as inadmissible (i.e. without examining its substance) when asylum seekers could have sought and, if eligible, received international protection in a non-EU country that is considered safe for them.

Frankly, this is untruthful. It is not serious to suggest that an asylum-seeker who made their way from Syria, Eritrea or Afghanistan to the European Union ‘could have sought’ international protection in Rwanda, a country many hundreds of miles from any route they would have taken. ‘Could have sought’ is not a rule in the legal text either. Trump-style dishonesty about migration and asylum should not be appearing in the output of the press office of an EU institution.

A Rwanda-clause – unlike the introduction of the transit clause in the ‘safe third country’ definition – has nothing to do with the route the asylum-seeker took, and everything to do with the destination country’s desire to dump the asylum-seeker in any State that will take them. This will inevitably become a key issue as regards the application of the law.

Adding the possibility of the EU asking non-EU countries for Rwanda treaties – as both the EP and the Council would like – does not change the game much. Although the EU has levers at its disposal to use the new transit clause – because its readmission treaties provide that the other parties must take back not only their own citizens, but also non-citizens who transited through their countries, and the EU’s visa code, visa waiver suspension rules, and soon trade policy laws, all sanction countries that do not comply – there are no such levers as regards Rwanda treaties. It remains to be seen what threats and bribes the EU and its Member States are willing to develop, and how easily the Rwandas of the world can be coerced or tempted by them.

 

European Public Prosecutor’s Office: the tension between supranationalism, sovereignty and legitimacy

 

Jacob Öberg, Professor of EU law, University of Southern Denmark

Photo credit: EPPO

The aim of this blog post is to summarise the key argument of a recent article published by the author in 50(6) 2025 European Law Review titled “The European Public Prosecutor's Office - supranationalism, sovereignty and legitimacy”.

For readers not acquainted with the European Public Prosecutor’s Office (“EPPO”) it could be briefly stated that the EPPO is a centralised European public prosecutor tasked with prosecuting and investigating crimes against the EU’s financial interests (as defined in the PIF Directive) following its mandate in Art 86 TFEU. EPPO which commenced its operation in June 2021, is undoubtedly the pinnacle to date of supranational criminal law in the history of European integration. It represents a significant achievement in terms of its potential for a fundamental system change for EU criminal policy, departing markedly from the conventional Member State-centric view that intergovernmental cooperation should remain the dominating principle of governance in this field (Öberg, 2021).

It is well-known that the EPPO was established in October 2017 through the EPPO Regulation on the basis of the new Art 86 , which provides the Council with a competence to ‘establish a European Public Prosecutor’s Office’ which shall ‘be responsible for investigating, prosecuting, and bringing to judgment … the perpetrators of, and accomplices in, offences against the Union’s financial interests’ and ‘exercise the functions of prosecutor in the competent courts of the Member States in relation to such offences’. The final EPPO Regulation had been preceded by politically protracted negotiations over four years, involving the highest number of official negotiation documents in the Council on criminal law to date (Eurocrim database) and a Yellow Card from national parliaments (Commission Communication, 2013).  Because of far-reaching objections from Member States to the creation of the office, the EPPO Regulation ultimately had to be adopted by means of an enhanced cooperation procedure involving 20 Member States under Art 86(1) 2nd para, TFEU.

In light of this brief account of the genesis of the EPPO, this blog offers a critical analysis of the evolution, structure and functioning of the EPPO on the basis of three theoretical frameworks: the theory of supranationalism, the concept of sovereignty and the critical approach of legitimacy. The first part of the analysis accounts for the current design of the EPPO along the supranational-intergovernmental (Stone Sweet and Sandholtz, 1997) spectrum based on the general literature on EU law and integration (Cappeletti, Seccombe and Weiler, 1986, Pescatore, 1974).  Secondly, the analysis proceeds to examine the governance of the EPPO in attempting to ascertain the extent to which Member States have been capable of maintaining control of its operation. Finally, we consider the EPPO from the perspective of legitimacy, with a specific focus on judicial review of the EPPO’s activities.

First, we reflect on the institutional structure of the EPPO. The key argument here is that the establishment and operationalisation of the EPPO marks a significant transformation from a ‘cooperative’ philosophy in EU criminal justice towards an integrated ‘supranational’ criminal justice system based on formal powers exercised by the EPPO (Monar, 2013). The EPPO Regulation nonetheless captures an intricate compromise between a supranational and intergovernmental conception of the EPPO (Schmeer, 2023). The central feature for the supranational characterisation is the creation of a European prosecutor with binding decision-making powers in respect of criminal investigations and prosecutions in the area of crimes against the EU’s financial interest (Art 86(2) TFEU and Art 13(1) of the EPPO Regulation), with jurisdiction transcending the territorial borders of the EU Member States (Art 23 of the EPPO Regulation). However, the complicated rules in the EPPO Regulation on the exercise of competence (Arts 25-27 of the EPPO Regulation), the removal of exclusive competence and the inclusion of a ‘national link’ in the EPPO’s governance structure (Art 13 of the EPPO Regulation) present limits to the ‘supranationalisation’ of the EPPO. A review of the first years of the EPPO’s activities nonetheless suggests that the body operates de facto as a highly supranational body without being restrained by the legal framework surrounding its operations. The limited evidence available indicates that the EPPO in practice defines its mandate broadly (both in terms of the PIF offences and in respect of ‘ancillary offences’) and that national law enforcement agencies acting on behalf of the EPPO act seemingly in a spirit of loyalty towards the EPPO’s interests (see Recital 69 of the EPPO Regulation). This lends some support to the contention that the EPPO – based on its legal powers and operational practice – is the most ‘supranational’ EU body created to date within the context of EU integration (Öberg, 2021).

The establishment of the EPPO, in conjunction with the adoption of the new PIF Directive, makes a compelling argument for holding that the EU appears to have adopted a ‘federal vision’ of criminal law, at least when it comes to protecting its financial interests (Herlin Karnell and Gomez-Jara, 2013). This development asks more fundamental questions about legitimacy and state sovereignty and if the EPPO stand as a role model for the creation of a ‘European criminal justice system’. It is important to observe that the EPPO exerts significant powers with severe implications for the fundamental freedoms of individuals and that the exercise of those powers also markedly encroaches on ‘core state powers’. Whilst there is a strong normative justification for conferring these enforcement powers to the EPPO (Öberg, 2024, ch 5), these powers need to be accompanied with strong fundamental rights safeguards both at national and EU level.

This brings us to the final observation which relates to the key fundamental rights challenges for a European ‘supranational’ prosecutor. While the establishment of the EPPO is a welcome step towards a ‘federalisation’ of EU criminal justice in this area, the EPPO cannot function effectively without some degree of harmonisation of national criminal procedures and national criminal laws. The recent case law of the Court (G.K. and others, Case C-281/22) highlights the implications of this incomplete centralisation of national criminal procedures which makes it more cumbersome for the EPPO to fulfil its task of combatting crimes against the EU’s financial interests. The ‘output’ (legitimacy) perspective aside, another central tenet of legitimacy for the EPPO is that there should be structures and mechanisms established to hold that body responsible and accountable for its actions. Not only must the legality of the EPPO’s decisions be subject to review by the Court of Justice, as follows from Art 263 TFEU, but national courts must also have the comprehensive ability to request preliminary rulings as per Art 267 TFEU. Therefore, the EPPO Regulation needs to be reformulated to make clear that the Court of Justice’s Treaty-based jurisdiction cannot be restricted by reference to secondary law. Furthermore, stronger common EU measures for protecting the rights of the defendant, along with effective safeguards that form the basis for the exercise of the EPPO’s powers, should be a central feature of a potential future amendment of the EPPO Regulation.

As suggested in the analysis, the Court can have an important role to play in this regard. First, the Court should expand its jurisdiction, following Art 47 of the Charter, to ensure robust judicial review of the actions of the EPPO, and by developing common standards constituting the basis for the EPPO’s operation through the autonomous interpretation of key provisions in the EPPO Regulation. Secondly, the Court can contribute to the development of a system of EU judicial remedies against actions undertaken by the EPPO in its operational activities. As demonstrated by G.K. and Others (Parquet européen) and EPPO v I.R.O. & F.J.L.R. (Case C‑292/23), the CJEU has taken a nuanced approach to judicial remedies, carefully balancing the need for an effective supranational system of criminal enforcement with judicial safeguards for defendants. To conclude, a supranational prosecutor such as the EPPO does not only need to have institutional structures, tools and resources to fight effectively against the EU’s financial interests effectively (output). It also needs a robust legal framework surrounding the EPPO’s operational action, combined with strong legal safeguards for individual defendants, is imperative for ensuring the (throughput) legitimacy of this new unique supranational prosecutor.