The Fiction of Non-Entry Meets the Fiction of Remaining: AG Emiliou in Sedrata

 

 

Andreina De Leo*, Post-doctoral researcher, Maastricht 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: Fred Romero, via Wikimedia commons

 

Introduction

 

On 23 April 2026, Advocate General Nicholas Emiliou delivered his Opinion in Sedrata, one of two pending preliminary references concerning the Italy–Albania Protocol. The Opinion marks a significant development in the ongoing proceedings before the Court of Justice of the European Union, addressing for the first time the question of whether asylum and return procedures may be conducted in a third country while remaining fully under the jurisdiction of a Member State.

 

While affirming that EU law continues to apply wherever a Member State exercises jurisdiction, and that the relocation of procedures outside EU territory does not, in itself, remove them from the scope of the acquis, the Opinion does not exclude, in principle, the possibility of delocalising asylum and return procedures to a third country, provided that all guarantees under EU law are fully respected.

 

This post examines the key elements of the Opinion and highlights the tensions that arise from its underlying logic.

 

EU Law Beyond Territory: Applicability and Conditional Compatibility in the AG Opinion

 

A central feature of the Opinion is its unequivocal rejection of a strictly territorial understanding of EU migration law. The AG finds that both the Return Directive and the Asylum Procedures Directive apply directly to the situations at issue, not merely by virtue of their incorporation into national law, but as a matter of EU law itself. In his view, the transfer of individuals to centers located in a third country cannot have the effect of rendering those directives inapplicable or of prejudicing the application of common standards, insofar as the procedures clearly fall within the material scope of EU law (see, in particular, paras. 34 and 48).

 

More broadly, the Opinion endorses a jurisdiction-based understanding of the scope of EU law, echoing principles familiar from international law. Where a Member State exercises authority over individuals, it remains bound to ensure compliance with all applicable EU rules, irrespective of the geographical location in which those activities are carried out. In this respect, the Opinion directly contradicts the position advanced by the European Commission during the hearing, which emphasized the territorial anchoring of the asylum acquis and denied its direct applicability outside the Union. Instead, the Advocate General affirms that EU law does not “switch off” when procedures are relocated beyond EU borders (para. 47).

 

Having established that EU law applies, the Opinion adopts a relatively concise approach to compatibility. As regards the Return Directive, the AG considers that nothing in its provisions expressly precludes the operation of detention facilities outside the territory of a Member State. In his view, the Directive regulates the grounds, duration, and conditions of detention, but remains silent on the geographical location of such facilities. From this silence, he infers that EU law does not, in principle, prohibit transfers to centers situated in a third country. The Opinion recognizes that such a scenario was likely not envisaged by the EU legislature. On this basis, and in the absence of explicit regulation, Member States retain a margin of discretion as to the organization of detention, including its possible extraterritorial dimension (paras 58-59, and 64).

 

At the same time, the Opinion draws a crucial conceptual distinction: the transfer to Albania cannot be qualified as a “return” or a “removal” within the meaning of the Directive. It neither brings the return procedure to an end nor constitutes the enforcement of a return decision. This clarification is significant, as it confirms that the Italy–Albania model is conceptually and legally distinct from the idea of “return hubs” currently discussed in the context of the proposed EU Return Regulation. In that framework, the transfer of a third-country national to a third State, other than the country of origin or habitual residence, would amount to the execution of a return decision directed towards that State, a possibility which is not permitted under the existing EU law. By contrast, the transfer to Albania is characterized as an intermediate step within an ongoing procedure: individuals remain fully under the authority and jurisdiction of the Member State, and the return decision is neither executed nor exhausted (paras 60-62).

 

A similar reasoning is adopted in relation to the Asylum Procedures Directive and the Reception ConditionsDirective. The AG considers that Article 9(1) of the Procedures Directive – which enshrines the right to remain on the territory of Member States pending the first instance asylum decision – does not, in principle, preclude the transfer of asylum seekers to a third country, provided that applicants are able to exercise their procedural rights effectively, including appearing before the competent authorities and benefiting from the guarantees laid down in EU law. While acknowledging that a literal reading might suggest that applicants must be brought back to the territory of the Member State once they lodge an application, the AG rejects such an interpretation when the provision is read in context and in light of its purpose. In his view, the notion of “remaining in the Member State” must be understood by reference to the definition in Article 2(p) of the Directive, which already includes border areas and transit zones, thus pointing towards a functional rather than strictly territorial understanding (paras 93-97).

 

On this basis, and supported by a functional and systemic reading, he argues that Member States cannot be regarded as precluded, as a matter of principle, from organizing asylum procedures in facilities located outside their territory, provided that those facilities fall under their jurisdiction and that EU guarantees are fully ensured in practice. The core objective of Article 9(1) is, in his view, to prevent removal before a final decision on the application has been taken and to safeguard the effectiveness of the right to seek asylum, rather than to guarantee physical presence on the territory of the Member State in all circumstances. In situations where those guarantees are maintained and no risk of refoulement arises, the relocation of procedures does not, in his view, undermine the Directive’s purpose. Accordingly, Emiliou concludes that Article 9(1) does not, in principle, preclude national legislation such as that at issue in the main proceedings (paras. 102–103).

 

The Functional Turn: “Magically turning non-EU territory into EU territory via some mystical legal alchemy” (Peers, here)

 

As argued elsewhere (see here, with Celoria, and here with Celoria and Ferri), I concur that EU law does not apply as a matter of voluntary choice, but follows from the fact that the procedures fall within its material scope, thereby ensuring the need to preserve the primacy and uniform interpretation of EU law, and to prevent unilateral circumvention of the acquis. Likewise, the transfer to Albania does not amount to a “return” or a “removal” within the meaning of the Return Directive, nor can it be assimilated to the concept of “return hubs”, as it does not constitute the enforcement of a return decision. It is also plausible that the silence of the EU legislature reflects the fact that such configurations were not contemplated at the time of adoption. However, beyond these points, the reasoning of the AG appears unconvincing, and is affected by internal inconsistencies.

 

The Opinion relies on a broad and functional interpretation of spatial concepts within EU law. The AG adopts an expansive reading of notions such as “territory,” “border,” and “transit zones,” suggesting that their meaning should be determined in light of function rather than physical location. This approach makes it possible to treat facilities located outside the Union as functionally equivalent to border or transit zones situated within Member State territory.

 

This is not entirely new: for instance, in the recent Danané (discussed, in this blog here), the Court, on the basis of an Opinion by the same AG, accepted that procedures traditionally associated with entry control may operate through a legal fiction of non-entry and be carried out in facilities located inland where national law designates them as “border” or “transit” settings. In that sense, the underlying logic of functional interpretation of geographical concepts is not unprecedented.

 

However, those cases remain anchored, in practice, within the geographical space of the Union, understood in a legally fragmented but still territorially contained manner. What is distinctive in the present context is the further conceptual step that is being contemplated: the extension of this legal fiction beyond the geographical borders of the Union itself. This marks a qualitative shift, as the “border” is no longer merely redefined within the territory of the Member State, but effectively displaced outside it.

 

This development raises significant concerns. By effectively allowing Member States to shape the spatial reach of EU procedures, it risks extending derogatory regimes beyond their traditionally circumscribed scope and weakening the safeguards attached to them. Moreover, the Opinion itself appears to expose tensions regarding the internal coherence and normative limits of this reasoning.

As mentioned, the interpretative ambiguity stemming from the absence of an explicit prohibition on relocating EU asylum and return procedures outside the Union is addressed through an expansive reading of “territory” and of “border” or “transit zones,” extending these concepts to extraterritorial settings. In other words, in the absence of a clear treaty or legislative definition clarifying that “territory” refers to the geographical territory of the Union rather than what a Member State unilaterally designates as such, the question arises as to whether Member States retain any discretion to determine the spatial scope of EU asylum law in this field.

 

EU institutions, including the Commission during the hearing, have consistently rejected such an approach, clarifying that asylum claims cannot be lodged or processed outside the Union, as this would require an extraterritorial application of EU law considered neither legally feasible nor politically acceptable. At most, Member States may transfer individuals to third countries in compliance with the principle of non-refoulement, but without retaining responsibility for the examination of asylum applications after disembarkation. The European Parliament’s Legal Service has adopted a similar position, recognizing that jurisdiction may, in certain circumstances, extend extraterritorially, while stressing that this does not transform such locations into EU territory for the purposes of asylum law. The institutional position thus converges on a key distinction: while jurisdiction may extend beyond borders in limited situations, the notion of “territory” under EU asylum law remains autonomous and confined to the geographical territory of the Union (for more info, see De Leo & Celoria, p. 604).

 

The AG introduces, however, a preliminary contextual remark noting that the agreement underlying the Italian legislation concerns Albania, a geographically close State to Italy, separated only by the Adriatic Sea, which is also an ECHR Contracting Party and an EU candidate country engaged in accession negotiations. These elements are treated as relevant factors in assessing the lawfulness of the scheme under EU law (paras 68-69). This is where a significant internal incoherence emerges.

 

If “territory” under EU asylum and migration law is not an autonomous legal concept, meaning that Member States are free to define its spatial reach at their discretion, then the assessment of compatibility with EU law in abstracto cannot, without contradiction, be made to depend on contextual variables such as geographical proximity or an ex ante assumption of compliance with fundamental rights. The attempt to anchor the analysis in Albania’s proximity and its presumed rights compliance therefore sits uneasily with the jurisdictional reading endorsed elsewhere in the Opinion. It effectively introduces an ad hoc relational criterion into what is otherwise presented as a non-autonomous legal concept under EU law, thereby blurring the distinction between a principled determination of the spatial scope of EU migration and asylum law and a pragmatic, case-by-case assessment of legality. Either the asylum and migration acquis is territorially bound to the Union’s geographical space, including only its legally constructed exceptions such as borders and transit zones, or it is not territorially constrained at all. In the latter scenario, the question would no longer concern abstract spatial limits but the concrete application of EU guarantees in individual cases, leaving national courts to assess compliance with EU law obligations without any role for EU law in further delimiting the spatial reach of “territory” itself.

 

The result is a conceptual inconsistency: what is framed as a question of legal scope, i.e. whether “territory” under EU law is autonomous or subject to Member State discretion, is made dependent on operative factors based on contextual and geopolitical considerations. These considerations, however, have no clear legal basis in determining the applicability of the acquis and end up suggesting, implicitly, a degree of autonomy in the concept of territory that the initial premise simultaneously denies.

 

This tension is further reinforced by the Advocate General’s acknowledgment that, while compatibility in abstracto may be preserved, the extraterritorial location of detention facilities inevitably generates practical and structural difficulties in ensuring full compliance with EU guarantees, thereby highlighting how relocation outside the Union introduces an inherent and additional layer of complexity in the effective protection of rights that would not arise within the territorial space of the Union.

 

More concretely, these difficulties are linked to structural factors such as geographical distance, the need to cross an international border to access the centers, and dependence on both third-country authorities and those operating within the centers for the enjoyment of key guarantees, including family visits and effective communication with legal counsel. The AG emphasizes that this does not, as such, imply a violation of EU law, but rather that Member States may be required to adopt specific organizational and logistical measures to ensure compliance with EU standards (para 80).

 

The limits of this reasoning become particularly visible in the AG’s own discussion of the requirement of “immediate” release under the Return Directive. In assessing the meaning of this guarantee, the Opinion accepts that the term “immediately” cannot be interpreted in an overly rigid or literal manner, but nevertheless acknowledges that national authorities must be able to organize, within a short timeframe, the practical modalities necessary to give effect to a judicial order ordering release, including transport by air or sea where required (para. 82). Even accepting this flexible interpretation, however, the Albanian model reveals the structural impossibility of equivalence introduced by extraterritorialisation. In a purely domestic setting, a finding that detention is unlawful triggers immediate release within the legal and physical space of the Member State, without any further material impediment to the restoration of liberty. By contrast, in the Albanian setting, even where a judicial authority orders release, the individual remains subject to continued physical constraint until the completion of a cross-border transfer back to Italy, thereby introducing an additional and legally significant phase of deprivation of liberty which is absent in a territorial context and stems solely from the extraterritorial location of the detention facility. This example illustrates, more broadly, that the issue is not confined to a single instance of unequal treatment. Rather, it reveals a pattern of structural frictions affecting a range of guarantees, i.e. access to legal assistance, the possibility of family visits, and the effectiveness of judicial protection, which, taken together, seriously undermine the claim of functional equivalence.

 

Seen in this light, the reasoning concerning the conditions for compatibility appears particularly unconvincing. The difficulty is not simply that individual guarantees may, in specific cases, be ineffectively implemented. It is that the extraterritorial design itself generates structural constraints that systematically hinder their effective exercise. These are not incidental or remediable deficiencies, but structural features of a system that operates outside the territorial and institutional framework in which the EU acquis was conceived. The issue, therefore, is not whether compliance can be demonstrated case by case, but whether a model that introduces additional layers of dependency, delay, and fragmentation can, as such, meet the standard of effectiveness and uniform applicability required by EU law, irrespective of any ad hoc organizational measures adopted by Member States. As such, this interpretation departs not only from a literal reading of the relevant provisions, but also from their teleological interpretation, which is precisely to ensure the effective and practical enjoyment of the rights they enshrine.

 

Furthermore, accepting the relocation of procedures outside the Union leaves unresolved a set of central issues concerning the Dublin system and mutual trust (a dimension addressed in detail in our previous analysis here), which the Opinion largely sidelines, merely noting that its applicability follows from that of the Asylum Procedures Directive and that its provisions must be interpreted coherently with it (para. 49). These questions therefore remain insufficiently developed in the reasoning, plausibly because they were not directly raised by the referring court, but are likely to be clarified in pending related proceedings (Joined cases C-706/25 Comeri and C-707/25 Sidilli), which more broadly concern whether the Protocol may adversely affect the proper functioning of internal EU rules, in potential breach of the Union’s exclusive competence in the light of the ERTA doctrine.

 

Conclusion

 

Overall, the Opinion of Advocate General Emiliou in Sedrata adopts an ambivalent approach to the legality of the de facto extraterritorial application of EU asylum and migration law.

 

On the one hand, it clearly affirms that EU law continues to apply wherever Member States exercise jurisdiction, thereby ruling out any attempt to circumvent obligations through geographical relocation. At the same time, it preserves the compatibility of the Protocol with EU law through a functional redefinition of the notion of “territory”. On the other hand, the delocalization of asylum and return procedures to third countries is accepted only under strict conditions: full compliance with EU standards, coupled with an uncertain reliance on contextual factors such as geographical proximity and an ex ante assumption of adequate fundamental rights protection based on formal international commitments. Yet the AG himself acknowledges that ensuring such compliance is significantly more demanding in an extraterritorial setting, while also leaving open a degree of Member State discretion in shaping the spatial understanding of “territory” and “border”.

Taken together, this gives rise to a structural tension and an internal incoherence that ultimately weakens the persuasiveness of the reasoning. From this perspective, coherence and effectiveness can in my view only be preserved by recognising that “territory”, for the purposes of EU asylum and return law, is an autonomous concept confined to the geographical territory of the Union, including its legally constructed exceptions such as border areas and transit zones. This implies, in practical terms, that both the examination of asylum applications under ordinary and accelerated border procedures and the detention phase under the Return Directive, prior to the enforcement of removal to the country of origin or to a third different country (should the proposed Return Regulation be adopted), must take place within the Union’s geographical space, irrespective of any unilateral qualification by Member States.

This view is grounded in the structural logic of the asylum and return acquis prior to removal, whose safeguards are designed to operate within the Union’s territorial framework. It is within this framework that supervision, enforceability, and effective protection are meant to be ensured. Accordingly, such guarantees cannot be transposed outside the Union through a functional extension of jurisdiction without altering their operation and effectiveness, even where formal complianc appears to be ensured.

Ultimately, it is now for the Court of Justice to decide whether this logic is to prevail, or should it be rewritten.

One Legal Fiction After Another: The Court of Justice judgment on the asylum border procedure in Joined Cases C‑50/24 to C‑56/24 (Danané)

 

 

Dr. Vasiliki Apatzidou, Associate Tutor at the University of London

 

Photo credit: Nicolas Economou, via Wikimedia Commons

 

Introduction

 

The Court’s judgment in the Danané case largely confirms the core argument made by Advocate General Emiliou in his Opinion: that the border procedure in EU asylum law (which provides for a fast-track assessment of asylum claims, with the implied legal fiction that the applicant has not entered the territory) is not exclusively attached to the physical border. In my earlier analysis of the Opinion, I argued that the Advocate General had moved towards a functional rather than pragmatic understanding of the ‘border’. According to him, procedures traditionally associated with entry control may not only take place in border facilities, but also in inland facilities if they are designated as such by national law. The Court of Justice has now broadly endorsed that logic by holding that EU law does not prohibit Member States from detaining asylum applicants during asylum border procedures in centres that are not geographically located at the border, and that such detention may continue in the same place after the expiry of the four-week period envisaged for the border procedure provided that a new legal basis for detention exists. In that sense, the original thesis still stands. However, the judgment also tries to contain the consequences of this argument through a focus on safeguards.

Summary of the Judgment

 

The joined cases arose from applications for international protection lodged at the Belgian border by third-country nationals who were refused entry and detained in the Caricole Transit Centre, a facility located inside Belgian territory but treated under national law as a place ‘at the border’ for the purposes of the border procedure. In Danané, the Court confirms that Article 43 of the recast asylum procedures Directive (APD), which sets out the current version of the border procedure, does not require border procedures to be implemented only in facilities physically located at the external border, since a border procedure may also be conducted in an inland detention centre that national law treats as a place ‘at the border’.

At the same time, once the four-week period in Article 43(2) recast APD expires, the case no longer remains within the border procedure and must continue under the regular procedure, even though the applicant may still be kept in the same facility if continued detention is separately justified under the detention grounds laid down in Directive 2013/33 (the reception conditions Directive, or RCD). The Court therefore accepts a dual classification of the same place of detention. However, this may happen only on condition that the applicant is informed of the change in legal status – it should be clear that he or she is now allowed to enter the territory – and that continued detention complies with the requirements of necessity, proportionality, individual assessment, and judicial review. It also makes clear that investigative steps already carried out during the border phase may still be relied on in the subsequent procedure, and that the examination of the application may be prioritized. Significantly, detention after the four-week limit cannot become automatic or systematic for all those who are channelled from the border to the regular procedure.

 

One Legal Fiction After Another

 

What makes Danané case especially important is that the Court relocates the decisive legal limit on the use of border procedures from geography to proper legal justification and safeguards. The judgment accepts that the same detention facility may first function as a place ‘at the border’ for the purposes of Article 43 recast APD and then, once the four-week period has expired, as a place ‘in the territory’, without any necessary change in the applicant’s physical surroundings. Instead, the Court tries to preserve a merely legal distinction between the two phases. Once the four-week period in Article 43 recast APD expires, the application no longer falls within the border procedure and his or her application will be examined under the other provisions of Directive 2013/32, while any continued detention may take place in the same facility, but must comply with the safeguards and requirements envisaged in Directive 2013/33. The Court also makes clear that detention cannot simply continue under the old border-procedure logic, that the applicant must be informed of the change in his or her legal situation, and that, where appropriate, the person must receive the document certifying their status as an asylum-seeker referred to in Article 6 of Directive 2013/33 or equivalent evidence, although in practice detained applicants often do not receive prompt documentation giving practical effect to this safeguard.

Yet the real problem is that this legal transition may leave the applicant’s position almost untouched in practice. If the same person remains in the same closed facility and under the same deprivation of liberty, the transition from border procedure to regular procedure risks becoming a legal fiction itself. That is where the judgment deserves a more critical reading. Formal entry into the territory may mean very little if it is not accompanied by adequate safeguards. The reception conditions Directive links applicant status to documentation, reception conditions, education for minors, and access to the labour market under the conditions laid down in the directive, yet the judgment says relatively little about how immediate and effective those rights must be in practice where the person remains detained. If an applicant still lacks prompt documentation, cannot effectively enjoy reception rights, and remains unable to exercise rights that normally follow from being an asylum seeker who has entered the territory and whose application is processed inside the territory under the regular procedure, then the shift from border to regular procedure serves primarily the asylum authorities, which in practice gain more time to decide while preserving continuity of control.

For that reason, continued detention after the expiry of the time limit foreseen in Article 43 recast APD must be treated with particular rigour. In this regard, it is very important that the Court itself turns from geography to safeguards, stressing that such detention must satisfy the ordinary standards of the recast RCD: necessity, proportionality, individual assessment, and one of the exhaustively listed grounds in Article 8 of the RCD, together with judicial review under its Article 9. If those requirements are not applied strictly, the transition endorsed in Danané risks turning the strict four-week time limit (extended to 12 weeks according to the Asylum Procedures Regulation, part of the EU Asylum Pact, which applies to asylum applications starting on 12 June) of the border procedure into a merely formal threshold.

This case matters beyond the Belgian transit centre and beyond the immediate interpretation of Article 43 recast APD. As my earlier analysis already suggested, the judgment fits a broader trajectory in EU asylum law in which the border is becoming less a fixed territorial line and more a shifting legal frontier that can be reproduced inside the territory. Read alongside the Pact texts, especially the Asylum Procedures Regulation, which more openly accommodates designated inland locations and expands the operational importance of border procedures, Danané looks less like an isolated ruling and more like a bridge toward a more internalised model of border procedures, in which procedures traditionally confined to the border and justified as exceptional may increasingly be reproduced inside the territory. For that very reason, safeguards become more important than ever, as the more flexible the geography of the border, the stricter the requirements governing detention, procedural guarantees, and the legal consequences of the expiry of time limits must be.

Conclusion

The deeper concern, then, is not merely that the Court has de-geographised the border. It is that, once the border is treated primarily as a legal status rather than a territorial threshold, ‘non-entry’ itself risks becoming a legal fiction capable of sustaining restrictive forms of asylum processing well inside national territory. The Court’s answer is that this is acceptable so long as the relevant guarantees, both in the asylum procedure and the detention safeguards, are preserved.

Whether that answer is convincing will depend entirely on practice. If detention after four weeks is truly exceptional and tightly reviewed, the distinction between border procedure and regular procedure may still make sense. But if applicants remain in materially unchanged detention while the determining authority simply acquires more time to decide, Danané will stand as a case of one legal fiction after another. Especially under the Pact, where screening and border procedures are further mainstreamed, the central question is whether safeguards are strong enough to prevent exception from becoming the ordinary mode of asylum governance.

 

 

 

 

The Aleb judgment on ‘safe third countries’ in asylum law: the CJEU’s answer to EU legislative amendments?

 

 

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.

Asylum Pact 2.0: the EU amends the rules on ‘safe third countries’ and ‘safe countries of origin’

 

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

Photo credit: Paula Reister, via Wikimedia Commons

Introduction

The EU has finally adopted two amendments to the EU’s asylum pact, entailing significantly more stringent rules as regards ‘safe countries’ of origin’ (adopted text here) and particularly ‘safe third countries’ (adopted text here).

This blog post summarises the new laws in their context, and then examines how much they impact upon the ability to claim asylum in the EU in practice – in particular the introduction 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.

Background

The history of the two ‘safe countries’ concepts in EU law goes 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 current 2013 asylum procedures Directive. The 2013 law will be replaced, as from 12 June 2026, by a new Regulation on asylum procedures, adopted in 2024 (the ‘2024 Regulation’, which I previously discussed here), part of the EU’s asylum pact. While the 2024 Regulation already makes the rules more stringent than in the 2013 Directive, the 2026 amendments now make those rules more restrictive again.

(For earlier blog posts on the 2026 amendments, see my analysis of the Commission proposal here, and my analysis of the Council and Parliament positions).

‘Safe countries of origin’

2024 Regulation

The 2024 asylum pact Regulation retains the definition of ‘safe country of origin’ from the 2013 Directive, referring to a list of human rights standards. As in the 2013 Directive, 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; and the principle remains a presumption, which the applicant can rebut in individual cases.

However, there are several changes. First of all, the rule will become mandatory for Member States, instead of an option. Secondly, 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’). Recent Court of Justice judgments have confirmed that such designations are not possible under the 2013 Directive (see here and here).

Thirdly, the 2024 Regulation makes it possible for the EU to adopt a common list of ‘safe countries of origin’, alongside Member States’ designation of such countries. Countries can be taken off the common EU 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; and 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.

Fourth, in place of a general reference to speeding up consideration of ‘accelerated’ cases (including ‘safe country of origin’ cases) in the 2013 Directive, there will be a three-month deadline to decide on such cases, although if an authority considers the asylum application to be too complex, it can default to the usual six-month deadline to decide on the merits. (This general deadline has not been amended, although the possibilities of extending it have been partly curtailed).

Fifth, a new ground for accelerated procedures, overlapping with the ‘safe third country’ ground, 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.’ It is also mandatory, and subject to the same deadlines.

Sixth, while the ‘safe country of origin’ concept remains a ground for applying special border procedures, as under the 2013 Directive, those border procedures have been changed. They are now mandatory in principle (for a specific number of asylum-seekers per Member State) and can now last for up to twelve weeks (including both the administrative and judicial phase of the case), instead of four weeks. Presumably the Court of Justice case law, requiring that asylum-seekers be released from detention (if that was the only ground for it) and admitted on to the territory at the end of the border procedure deadline, still applies. Unaccompanied minors will be exempt from the revised border procedure, unless they are ‘national security’ or ‘public order’ risks.

Finally, it will be mandatory, instead of optional, for Member States to deny automatic suspensive effect to appeals in ‘safe country of origin’ cases, and several safeguards related to the lack of automatic suspensive effect are dropped. However, it will still be possible for asylum-seekers to ask a court for permission to stay during the appeal. On appeals, there is nothing to suggest that the case law of the Court of Justice on the 2013 Directive (Alace) has been overturned. In that case, the Court said that: a Member State’s designation of a ‘safe country of origin’ has to 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.

2026 amendments

First of all, the new Regulation allows (as an option) the early application (ie before June 2026, from the day after publication of the new law in the EU Official Journal) of the possibility of partial designation of a country as a ‘safe country of origin’ (either geographically or for certain groups of asylum-seekers), as well as the new ‘20% recognition rate’ rule. In particular, this is intended to allow for the earlier application of the Italy/Albania arrangements on processing asylum-seekers on Albanian territory, which were in effect frustrated by the Court of Justice’s interpretation of the current Directive. Of course, other Member States might have a use for earlier application of the rules too.

Furthermore, the 2026 amendments similarly allow for early application of the partial designation of countries as ‘safe third countries’.

The early application of the 2024 version of these rules can also be applied at the borders, in the current version of the border procedure. This means that, among other things, the 2024 exclusion of most unaccompanied minors from the border procedure will not apply until the 2024 Regulation fully enters into force in June.

Secondly, the 2026 Regulation sets out a common EU list of ‘safe countries of origin’, as from the entry into force of the Pact: 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 are, however, subject to special rules: they cannot be regarded as ‘safe’ if they have been subjected to EU foreign policy sanctions due to their actions ‘affecting fundamental rights and freedoms that are relevant to’ the definition of ‘safe country of origin’; or if their recognition rate is above 20% at first instance; or if there is a ‘serious threat to a civilian’s life or person by reason of indiscriminate violence’ in that country (which is an amended version of one of the grounds for ‘subsidiary protection’ in EU law; the requirement that the threat also be ‘individual’ does not apply in this context). The preamble draws the obvious conclusion that ‘Member States should not apply the concept of safe country of origin to applicants from a candidate country during the period in which the circumstances provided for by this Regulation persist.’ 

It appears that in place of the usual rules on suspending a country from the list, the accession candidates will be removed following a notification from the Commission to the Council, European Parliament and Member States. However, to trigger the ‘serious threat’ ground for removal from the list, the Commission will need approval from the Council first. Presumably the default voting rule of qualified majority in the Council will apply. According to the preamble, the Council’s role is necessary ‘in view of the potential implications for the external relations of the Union and the Member States’ in this scenario.

Several of these points (dropping the ‘individual’ criterion, qualifying the sanctions point, and creating a role for the Council) were changes from the Commission’s original proposal, made at the Council’s behest. The European Parliament’s proposed amendment (removing a country from the list if the temporary protection Directive applied – obviously referring to Ukraine) was not agreed, although the final version of the preamble says that the application (or termination) of temporary protection under that Directive should be taken into account when assessing whether a ‘serious threat’ exists. Moreover, the Parliament’s proposal to use delegated acts (ie the usual process) to remove candidate countries from the common list was not agreed either.

The Regulation also now provides for the possibility (again, at the Council’s behest) that a country on the common EU list can be de-listed in part, ie using the exceptions for parts of a country or groups of people which Member States can use.

Finally, at the behest of the Parliament, a clause in the preamble to the 2026 amendment appears to hint at limiting judicial review in individual cases:

[a ‘safe country of origin’] designation reflects the general situation in that country and is not affected by the individual circumstances, which may only be assessed for the purposes of determining whether the concept of safe country of origin should exceptionally not be applied in a specific case. Therefore, in the context of national judicial review, the detailed evidence regarding an applicant’s individual situation justifying the applicability of the concept of safe country of origin should be the main purpose of that assessment. In accordance with the Treaties, the Court of Justice of the European Union is competent to rule on any doubts on the validity of a designation of a third country as a safe country of origin at Union level.

This is slightly more equivocal than the EP’s proposed amendment on this point, which had read:

…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’

2024 Regulation

The 2024 Regulation (before the 2026 amendments) makes a number of changes to the rules on this issue as compared to the 2013 Directive. First of all, it will be simpler to designate a country as a ‘safe third country’ in one respect: there need not be an opportunity to request and receive Refugee Convention refugee status there. Instead, 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’.  

Secondly, as with ‘safe countries of origin’, it will now be 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.

Thirdly, there will be new safeguards for unaccompanied minors: a ‘best interests’ test and a prior assurance that the non-EU country will take them in charge and give them immediate access to effective protection.

Fourth, 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.

Fifth, there will also now be 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.

Sixth, the revised border procedure in the Regulation (described above), will also apply to ‘safe third country’ cases.

Several elements have not changed, though. It is still necessary that, in the country concerned, there is: 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); and 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’.

Also, the principle will still only be optional for Member States; and the rule still cannot be applied unless the asylum-seeker will be admitted to the country concerned. Also, the rule can only apply if the asylum-seeker has a ‘connection’ with the other country, ‘on the basis of which it would be reasonable’ for the asylum-seeker to go there; it must still be possible for the asylum-seeker to challenge the alleged safety of the country concerned for them, as well as their supposed connection with it. Finally, as before, appeals in ‘safe third country’ cases will still have automatic suspensive effect.

2026 amendments

First and foremost, the new Regulation (which will apply from June, along with the rest of the asylum pact) has dropped the requirement that the ‘safe third country’ principle can only be applied in the event of a ‘connection’ to the supposedly ‘safe’ third country. There are further explanations of that principle in the preamble:  

Member States should, while fully respecting the parameters set out in the case law of the Court of Justice of the European Union, be able to apply the concept of safe third country 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 third country, where the applicant has settled or stayed in that third country, or where the applicant has linguistic, cultural or other similar ties with that third country.

It will now also be possible to apply the ‘safe third country’ principle in two other cases. First, it can be applied where the asylum seeker had transited through that country on the way to the EU, as further explained in the preamble:

…transit through a third country could include a 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 that applicant has had the possibility to request effective protection from the authorities of the third country concerned.

Secondly, the ‘safe third country’ principle can now apply to a Rwanda-type deal with a country that the asylum-seeker had neither a connection with nor transited through: ‘there is an agreement or an arrangement [with the non-EU country concerned]…requiring the examination of the merits of any requests for effective protection made in the third country concerned by applicants subject to that agreement or arrangement’.

Such an agreement or arrangement can either be negotiated by the EU or one or more Member States; the Member States might even negotiate it alongside non-EU countries (say the UK, or the US?). There are detailed rules on how any EU negotiations relate to Member State treaties on the issue.

However, the new ‘Rwanda’ criterion will not apply to unaccompanied minors, although the Parliament had argued that it should apply to children who were security threats. The new transit criterion will apply to unaccompanied minors though, although the current safeguards for them in ‘connection’ cases (best interests of the child, certain assurances from the country concerned) in the 2024 Regulation will apply to ‘transit’ cases too. The preamble also refers to considering the family unity principle.

Finally, the new Regulation has altered the rules on appeals, so that there will no longer be automatic suspensive effect in ‘safe third country’ appeals, except for unaccompanied minors subject to the border procedure. The amendments also (at the Council’s behest) drop automatic suspensive effect in cases where the asylum seeker has international protection from another Member State, even though 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. Nevertheless, it will still be possible to request a court to grant suspensive effect.

Other features of the ‘safe third country’ principle have not changed: it will remain optional for Member States, and the rules will not change as regards criteria defining the ‘safety’ of a country as such, the applicable deadlines, the related rules on the border procedure, or the requirement that Member States must consider the merits if the third country concerned refuses to admit the asylum seeker. Nor does it use the power to adopt a common EU list of ‘safe third countries’.

 

Assessment

The ‘safe country of origin’ amendments in particular are partly aimed at overturning case law of the CJEU, not only as regards allowing geographical and group exceptions from the principle, but also (on a more lasting basis, given that those exceptions would be allowed from June anyway) as regards limiting national judicial review of the national lists of such countries.

However, this apparent attempt to limit judicial review would, if the provision is interpreted as such, breach 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.

In any event, the litigation has not stood still. There are further cases referred from the Italian courts as regards aspects of the Italy/Albania arrangements, including Sedrata on detention issues, and Comeri (fast-tracked by the Court) on the external relations aspects of such arrangements. These pending cases may limit the possible application of the Italy/Albania agreement in the meantime, although it remains to be seen whether the changes to the law (including the 2024 version of the reception directive), might be relevant to the issues in these disputes post June. Of course, further legal challenges can be expected as the 2024 Regulation is partly rolled out early on the basis of the 2026 amendments, and applied fully from June.

The removal of automatic suspensive effect from ‘safe third country’ cases, while widening the scope of such cases in principle, may mean that it is harder to obtain an effective remedy in such cases – although the possibility of requesting suspensive effect from the national court will now become crucial. The same will be true of cases where a Member State is allegedly breaching the Charter rights of recognised refugees – an issue separate from ‘safe third (non-EU) countries’ where the legislators decided to carry out a ‘drive-by hit’ on appeal rights nonetheless.

As for the ‘Rwanda clause’ itself, at first sight it leaves open the possibility that (apart from unaccompanied minors) literally everyone who applies for asylum in the EU could be removed to such a country. In practice, though, this depends upon there being enough countries willing to admit the asylum-seekers. The EU’s current and planned leverage (via the visa code, recent visa waiver suspension rules, agreed amendments to trade policy laws, proposed changes to development aid and planned further changes to the visa code) relates to countries taking back their own citizens and those who transited through them. While the transit point is relevant to the widening of the ‘safe third country’ concept to include transit states, it does not cover the introduction of the Rwanda clause.

Even for those who are subject to the wider application of the ‘safe third country’ principle after the 2026 amendments, it will still be necessary to show that they will be admitted to the country concerned, and still possible to argue that the country is ‘unsafe’ in their particular situation. Moreover, it could be argued that – in parallel with the CJEU case law on challenging the listing of ‘safe countries of origin’, quoted from above – the EU Charter of Rights necessarily confers the possibility of challenging the listing of the country in question itself, comparable to the UK Supreme Court ruling that Rwanda was ‘unsafe’ as such. And since the EU Charter has a higher legal status than EU legislation, it would not be possible for the EU or its Member States to overturn such a ruling merely by passing legislation – unlike the ‘Safety’ of Rwanda Act passed under the previous UK government.