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.

Testing the Applicability of EU Law Abroad: The Italy–Albania Protocol in the Comeri, Sidilli, and Sedrata Hearings

 

Eleonora Celoria, Post-doctoral researcher, FIERI

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

Marcella Ferri, Research Fellow, Florence University

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

Photo credit: Jorge Franganillo, via Wikimedia Commons

Introduction

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

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

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

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

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

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

Outline of the post

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

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

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

The Italian government

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

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

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

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

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

The European Commission

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Conclusion

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

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

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

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

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

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

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

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

The judgment of the Grand Chamber of the Court of Justice of the European Union on the Italy-Albania Protocol

 

 

 

Matteo Zamboni (human rights lawyer working between Italy and the UK. He is a partner to the Immigration Law and Policy clinic of Goldsmiths, University of London)

Photo credit: Shëngjin Port in Albania, the location of one of the two detention centres built by Italy © Albinfo, CC BY 4.0 via Wikimedia Commons

 

Introduction

 

On 1 August 2025, the Grand Chamber of the Court of Justice of the European Union (CJEU) published its judgment in joined cases C-758 and 759/24 (Alace and Canpelli) regarding the Italy-Albania Protocol, and more specifically the designation of safe countries of origin under Article 37 of EU Directive 2013/32 (the currently applicable asylum procedures Directive). In previous blogs on the public hearing of 25 February 2025 and the opinion of the Advocate General (AG) of 10 April 2025, I gave some information on the context of the case and its high political stakes at Italian and European level. Indeed, the plan is not only critical to the political capital of the Italian government but also admired by foreign heads of government, including the UK Prime Minister, and taken as an example (or a ‘general rehearsal’) by the President of the EU Commission.

 

On this latter point, it is probably worth repeating that, while the present case was pending before the CJEU, the EU Commission:

 

-          changed its position on whether, under Directive 2013/32, Member States have the power to designate a country as a ‘safe’ country of origin with the exclusion of ‘clearly identifiable categories of persons’, by first denying this possibility in its written pleadings of (where it was argued that ‘Directive 2013/32 does not allow for the designation of a third country as a safe country of origin where a significant number of persons belonging to the same category are systematically persecuted or face a risk of suffering serious harm […]’) and then admitting in the oral arguments made at the hearing (to the surprise of the President of the Grand Chamber);

-          adopted a proposal on the idea of “return hubs” located outside the European Union, with a move that mirrors the legislative amendment enacted by the Italian government while awaiting the decision of the Grand Chamber (with decree-law no 37 of 28 March 2025) to use one of the two centres built in Albania (the one in Gjader) as a return hub to detain individuals whose claims to international protection had been rejected and who are awaiting deportation; and

-          published a proposal for amendments to Regulation 2024/1348 (the asylum procedures Regulation, applicable from June 2026) which would, among other things: (i) create a European Union-level list of ‘safe countries of origin’, including Bangladesh, the country of origin of the real individuals behind the fictitious names of Alace and Canpelli; (ii) bring forward the possibility, for Member States, to apply the new provisions allowing to designate a country of origin as ‘safe’ in spite of the existence of ‘specific regions’ and/or ‘clearly identifiable categories of individuals’ for which the country is, in fact, not safe.

 

The extreme salience and significance of the case is equally witnessed by the intervention of 16 Member States that, like the Commission, endorsed the position of the Italian Government. It is no surprise, then, that the news that the judgment of the Grand Chamber granted all of the arguments made by the defence on behalf of the asylum seekers concerned was met by a strong reaction in the media and from the Government. As a matter of fact, in a long tweet posted on the day of the judgment, the Italian President of the Council of Ministers, Giorgia Meloni, accused the Court of Justice to ‘claim[…] powers that do not belong to it’ by ‘decid[ing] to delegate to any national judge the decision not on individual cases, but on the part of migration policy relating to the repatriation and expulsion of illegal immigrants’.

 

The tweet (which is very long and available here) betrays the disappointment for the considerable difficulties and drawbacks in the enforcement of the scheme. As denounced by the Italian watchdog for the region of Lazio, who visited the centres on 30 July 2025 with his counterpart for the city of Rome, just before the judgment, the centres accommodated an ‘extremely limited number of people’ (27) which could easily be detained in the available places in the Italian structure. Against this background, one could think that the CJEU judgment would mark the end of the scheme; but this would probably be too optimistic. In fact, the CJEU did neither ban the conduct of asylum procedures in Albania, nor delved deep into the many human rights concerns inherent in the extra-territorial processing of asylum claims, which have been already highlighted in the legal literature, and to some extent already referred to the CJEU for further consideration with an order of the Italian Court of Cassation of 29 May 2025-20 June 2025 (no 23105).

 

So, while this ruling is certainly ‘another blow to a key aspect of the Italian government's migration policy’, it will most likely not be the final one. Instead, as reported in the press, from the day following Meloni’s reaction on Twitter some of her ministers are already changing their tune, saying that the CJEU judgment is actually good, because it explains to the Italian government the changes to be enacted to (finally!) activate the centres in Albania.

 

Summary of the judgment

 

The judgment focusses on the designation of safe countries of origin (SCOs), which, as explained by Judge Jürimäe, bears far-reaching consequences for asylum seekers. It triggers the application of the accelerated examination procedure and the operation of the (rebuttable) presumption that no protection is needed. The question for the Grand Chamber of the CJEU, then, was which national authority has the final say over the designation of a country as ‘safe’ under Article 37 of Directive 2013/32 and whether a country can be considered as ‘safe’ in spite of the existence of exceptions for vulnerable categories. These questions are interesting from the point of view of EU law, since, as it has been noted, the CJEU has dealt with the issue only in one prior case; i.e., the judgment of 4 October 2024 in case C-406/22 (CV), regarding the designation of a country as ‘safe’ despite the existence of territorial exception (more specifically, the case was concerned with the designation of Moldova as a SCO by the Czech Republic with the exception of the region of Transnistria). Therefore, from a general perspective, a new (and clarificatory) intervention from the Luxembourg judges was much needed.

 

However, in the specific circumstances of the Italy-Albania Protocol, the question is even more critical, as it is precisely the designation of the country of origin of an asylum seeker as ‘safe’ that allows the deportation to the centres in Shengjin and Gjader  - and ultimately the operation of the whole scheme.

 

Under the applicable legislation (the Protocol of 6 November 2023, as integrated by the Standard Operation Procedures of the Ministry of the Interior, and the ratification law no 14 of 21 February 2024), prior to the amendments enacted by decree-law no 37 of 28 March 2025, passed into law no 75 of 3 May 2025, the requirements for deportation and detention into the centres in Albania were:

 

-          the fact that a ‘migrant’, as defined by Article 1(d) of the Protocol, is intercepted by the Italian navy on the high sea (see Article 4(4) of the Protocol; paras 4-6 of the SOPs; Article 3 (2) of law no. 14/2024, which makes explicit reference to ‘persons taken aboard Italian authority vessels outside the territorial waters of the Republic or other Member States of the European Union, including as a result of rescue operations’);

-          the consideration that the person concerned does not belong to one of the vulnerable categories established by law (i.e., women, non-accompanied minors, individuals affected by clear pathological conditions, elder people - see para 1 of the SOPs); and

-          the assessment that the asylum seeker comes from a country designated as a SCO (see para 6 of the SOPs). 

 

In terms of the domestic legal system, the latter requirement stems from Article 3(3) law no 14/2024, which equalises the centres in Albania to the border and transit areas provided for by legislative decree no 25 of 28 January 2008. This is the legislation that Italy passed to implement Directive 2005/85 on minimum standards on procedures in Member States for granting and withdrawing refugee status (the previous asylum procedures Directive), and that was then amended and integrated to transpose Directive 2013/32 and further EU legal instruments regarding asylum procedures. Among many other things, this legislation established a list of SCOs (at Article 2-bis, as amended by Article 1 of decree-law no. 158 of 23 October 2024) and stated that the request for international protection filed by individuals coming from one of such countries can be dealt with under the accelerated procedure (see Article 28-bis).

 

Moreover, Article 3(4) law no. 14/2024 states that the centres in Albania are to be considered as equivalent to the reception centres established by Article 10-ter(1) of legislative decree no. 286 of 25 July 1998; i.e., the so-called ‘special crisis centres’ in which are detained the ‘foreign nationals who are tracked down while crossing internal or external borders illegally, or who arrive on national territory following rescue operations at sea’. In terms of EU law, basically Italy relied on the designation of SCOs under Article 37 Directive 2013/32 to trigger the accelerated procedure under Article 31(8)(b) of that Directive and considered that the examination of the asylum request from the centres in Albania would be equal to the border procedure provided by Article 43 of the same directive. Therefore, the designation of the country of origin of the migrants concerned as a SCO is the necessary preliminary point for the operation of the entire scheme, including deportation to and detention in the centres in Albania.

 

The questions referred to the CJEU

 

For the sake of clarity, it is worth restating that the questions referred to the CJEU by the Tribunal of Rome in November 2024 concerned whether EU law, and in particular Articles 36, 37, and 39 of Directive 2013/32, interpreted in the light of Article 47 of the EU Charter on Fundamental Rights (‘the Charter’),

-          prevents Member States from designating SCOs by means of legislative instruments (acts of Parliament);

-          requires national legislation to publish the sources relied on for the designation of a specific country as a SCO;

-          allows national judges called to review the designation of SCOs to use information from sources other than those referred to in directive 2013/32; and

-          precludes a non-EU country from being designated as a SCO where there are categories of people for whom the substantive conditions for such a designation laid down in Annex I to directive 2013/32 are not met.

 

The answers of the Grand Chamber

 

At the outset, the 15 Judges assessed the admissibility of the request for referral, reminding that, according to settled case-law, ‘questions on the interpretation of EU law referred by a national court […] enjoy a presumption of relevance’ and can be refused ‘only [i] where it is quite obvious that the interpretation of EU law sought bears no relation to the actual facts of the main action or its purpose, [ii] where the problem is hypothetical, or [iii] where the Court does not have before it the factual or legal material necessary to give a useful answer to the questions submitted to it’. Clearly, in the mind of the Grand Chamber, none of these instances applied to the case at hand, which, on the contrary, raises questions of interpretation of Directive 2013/32 highly relevant for the solution of the case (see judgment § 38-41 with reference to the judgment of 19 December 2024 in cases C-185 and 189/24 (Tudmur) § 26)

 

The first question and the issue of disapplication

 

On the first question, the Grand Chamber followed the argument advanced by all the parties at the hearing, including the defence, and by the AG in his opinion, stating that the term “legislation” contained in Article 37(1) of Directive 2013/32 is to be interpreted in a wide manner; and that neither Article 37 nor other provisions of Directive 2013/32 sets out the specific national authority competent for the designation or the specific legislative instrument to be used for that purpose (ibid § 56 and 59-60). Finally, the Court reminded that, under Article 288(3) TFEU, Member States enjoy a margin of discretion when implementing directives (ibid § 61). So, the Court concluded that nothing prevents a Member State from issuing a list of SCOs by means of an act of Parliament.

 

At the same time, albeit acknowledging that, as stressed by the Italian Government and by all the intervening Member States, EU law leaves discretion to States when issuing the list of SCOs, the Court clarified that said discretion does neither affect (i) ‘the obligation […] to adopt all the measures necessary to ensure that the directive concerned is fully effective”, nor (ii) ‘the duty of the national judge to give full effect to the provisions of Directive 2013/32, including by dis-applying as required, of its own motion, any national rule, even if adopted subsequently, which is contrary to a provision of EU law with direct effect, without it having to request or await the prior setting aside of that national rule or practice by legislative or other constitutional means’ (ibid § 62 with reference to the judgments of 10 April 1984 in case C-14/83 (von Colson and Kaman) § 15 and of 31 March 2022 in case C-472/20 (Lombard Lízing) § 53; § 63 with reference to the judgments of 9 March 1987 in case C-106/77 (Simmental) § 21 and 24, and of 28 January 2025 in case C-253/23 (ASG 2) § 90).  

 

This argument was compounded by the reference to the right to an effective remedy and to a fair trial, enshrined in Article 47 of the Charter. Indeed, the Court quoted from the precedent of case C-406/22 and held that

 

‘Article 46(3) of Directive 2013/32, read in the light of Article 47 of the Charter, must be interpreted as meaning that, where an action is brought before a court or tribunal against a decision rejecting 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, even if that failure is not expressly relied on in support of that action’ (ibid § 66 with reference to case C-406/22 § 98)

 

On this basis, the Grand Chamber concluded that the fact that a Member State decided to designate SCOs by means of an act of Parliament, while not in itself contrary to Article 37 of Directive 2013/32 or any other EU law provision, cannot prevent the national judge to “check” (“controllare”) the designation, or “subject [it] to judicial control” (“oggetto di un controllo giurisdizionale”), on the basis of the substantive requirements set out in Annex I to the directive (ibid § 66 and 67 for the answer to the question).

 

The second and the third question on the right to access to the sources used for the designation of SCOs

 

As regards the second and the third question, the Court recognised that, while it lists a series of sources that need to be taken into consideration (namely, ‘information from other Member States, EASO, UNHCR, the Council of Europe and other relevant international organisations’), Article 37(3) of Directive 2013/32 does not expressly set out that the national authorities are under the obligation to publish and make accessible the sources used for the designation of SCOs (ibid § 70-71). Still, the Court followed the ‘systemic’ approach indicated by the AG in his opinion and applied a three-fold argument to reach the conclusion that (i) ‘the Member State designating a third country as a safe country of origin must ensure sufficient and adequate access to the sources of information referred to in Article 37(3) of [the] Directive’ and (ii) ‘the national judge […] may, if it verifies, even incidentally, whether such designation complies with the substantive conditions for such designation set out in Annex I to that Directive, take into account the information it has gathered, provided that, on the one hand, it ensures the reliability of that information and, on the other hand, it guarantees the parties concerned the right to be heard’ (ibid § 88).

In the first place, the Grand Chamber noted that the designation of a country as ‘safe’ triggers the operation of the presumption of sufficient protection in the country of origin of an asylum seeker. However, in order to guarantee the right to challenge and rebut the presumption, the asylum seeker must be put in the condition of knowing the reasons for the designation, including the sources employed by the national authorities’ (ibid § 72-73).

 

In the second place, the Judges underscored that Article 12(1)(d) of Directive 2013/32, read in combination with Article 10(3)(b), states that during the procedure regarding the examination of their claim, asylum seekers have a right to access ‘precise and up-to-date information […] from various sources, such as EASO and UNHCR and relevant international human rights organisations, as to the general situation prevailing in [their] countries of origin’. It, then, concluded that this information is basically the same as that referred to in Article 37(3). Moreover, under Article 10(4) of the directive, the national judicial authorities hearing an appeal against a decision of refusal of protection have access to the same information (ibid § 74-75).

 

In the third place, the CJEU relied - once again - on the right to fair trial, guaranteed by Article 47 of the Charter, as a guide in the interpretation of Article 46 of Directive 2013/32, which set out the right to an effective remedy, reminding that, as it had already been stated in case C-406/22 and other precedents, the right to effective judicial protection resulting from the combination of these two provisions ‘is sufficient in itself and does not need to be made more specific by provisions of EU or national law in order to confer on individuals a right which they may rely on as such’ (ibid § 77 with reference to case C-406/22 § 86 and case law therein cited). That said, the Court remarked that effective judicial control presupposes that both the asylum seeker and the judicial authority must have full knowledge of the grounds of the decision of refusal (ibid § 78 with reference to the judgments of 4 June 2013 in case C-300/11 (ZZ) § 53 and of 29 July 2024, C-185/23 (protectus) § 79).

 

Yet, where a request for protection is denied as manifestly ill-founded since an asylum seeker comes from a ‘safe’ country, the reasons for the refusal overlap with the grounds to hold that the said country is indeed to be designated as ‘safe’ under Articles 36 and 37 and Annex I of Directive 2013/32 (ibid § 79). The ensuing conclusion that the asylum seeker and the national judge must have access to the sources and information relied on by the national authorities for the purpose of designating a country as ‘safe’ was further compounded by the literal interpretation of Article 46 of the Directive in the part that state that ‘Member States shall ensure that an effective remedy provides for a full and ex nunc examination of both facts and points of law, including, where applicable, an examination of the international protection needs pursuant to Directive 2011/95/EU […]’.

 

In this regard, with an argument that it had already employed in case C-406/22, the Court stressed that the expression ‘ex nunc’ indicates that the examination of the judicial authority has to include ‘new elements intervening after the adoption of the decision of refusal [of protection]’; that the adjective ‘complete’ means that the national judge has to examine ‘all the elements [that] […] have, or ought to have, been taken into consideration’ in the decision of dismissal of the claim; and that the clause ‘where applicable’ highlights that the process of judicial review ‘does not necessarily have to focus on the […] merits of the request for international protection’, rather ‘it could be concerned with the procedural aspects [of the claim] […] including the designation of a third country as a safe country of origin’ (ibid § 81-84 with reference to case C-406/22 § 87-91).  

 

The fourth question on the “personal exception” to the designation of SOCs

 

Finally, the Court held that Article 37 and Annex I of Directive 2013/32 prevent a Member State from designating a country as a SCO where the substantial conditions set out in Annex I to the directive are not satisfied for certain categories of persons. The conclusion ensues from an analysis of the letter of Article 37; the context of that provision in the framework of Directive 2013/32; and the objective of the EU legislation (ibid § 91).

 

To begin with, the Grand Chamber noted that nothing in the text of Article 37 suggests the terms ‘countries’ and ‘third countries’ may be interpreted as regarding just a portion of the population, even if this portion would be the majority (ibid § 92). This was confirmed by an analysis of the context of the directive, and specifically of the “substantive criteria” listed in Annex I. Here, the CJEU disagreed with the argument put forward by the AG and held that, in spite of ‘semantic differences’ in the official translations, the adverbs ‘generally’ and ‘consistently’ (in the English version of the directive) refer to a notion of ‘invariability’ (“invariabilità” in the Italian version of the judgment, ibid § 93-96). Moreover, the Court repeated that, as it had already stated in case C-406/22, all clauses of derogation and exceptional provisions need to be interpreted in a restrictive manner (ibid § 100).

 

It is perhaps worth noting that, in giving the interpretation of Article 37 in accordance with the literal and context-based approach, the Grand Chamber does not reference case C-406/22, although the same reasoning features in that judgment at §§ 65-71. On the other hand, case C-406/22 is explicitly quoted in the argument regarding the interpretation of Article 37 based on the objectives of Directive 2013/32. Indeed, §§ 101-106 of the Alace and Canpelli judgment replicate §§ 77-82 of case C-406/22. In both judgments, the Court stated, in essence, that the purpose of Directive 2013/32 is to ensure that, even under the accelerated procedure, asylum claims receive ‘adequate’ and ‘complete’ examination, in compliance with ‘basic principles and guarantees’, and that the presumption of safety of a country of origin remains ‘rebuttable’. In this context, the Court concluded that, when enacting Directive 2013/32, the EU legislature exercised its ‘discretion’ in order to ‘strike a fair balance’ between the need to speed up the processing of asylum requests and the duty to ensure that these are given full and fair examination. In the framework of this balancing exercise, the EU legislature decided ‘not [to] provide […] for the option for Member States to exclude certain categories of persons for the purposes of such designation’. Falling within the ‘prerogatives’ of the EU legislature, this choice of the balance point cannot be questioned by Member States.

 

The same applies to the decision as to the date of entry into force of the new Regulation 2024/1348, allowing, at Article 61(2), for the designation as a ‘safe country of origin’ with exceptions for specific parts of its territory or clearly identifiable categories of persons. Once again, the Grand Chamber held that the EU Legislature decided that the new legislation will only come into effect in June 2026 (ibid § 101-106).

 

As anticipated, this reasoning is exactly the same as the one developed in case C-406/22 concerning the so-called “territorial exception”. The only difference is that the Alace and Canpelli judgment contains (at § 107) a further reference to the proposal of the Commission entailing the amendments to Regulation 2024/1348 and the anticipation of the possibility, for Member States, to apply the new provisions regarding the designation of SOCs to ‘as soon as possible before June 2026’. Even this initiative is seen as an example of the exercise of legislative prerogatives by EU institutions, which, according to the CJEU, further reinforces the conclusion that – as the legislation currently stands under the realm of Directive 2013/32 – EU law does not allow for group exceptions in the designation of SCOs.

 

Conclusion

 

As noted, the CJEU gave priority to the case of Alace and Canpelli and suspended the examination of all the other questions referred by Italian judges, including on the designation of SCOs. Yet, it is not certain that the judgment of 1 August will clear all doubts surrounding the issue, and that, as a consequence, it would allow national judges to resolve the cases before them independent from further guidance from the Court.

 

This seems to be particularly true as far as the issue of disapplication is concerned. In fact, the Grand Chamber does not seem to answer in full to the argument, put forward by the Italian government, that the effective remedy available to the concerned asylum seeker to question, in general terms, the designation of a his or her country of origin as ‘safe’ is a challenge of constitutionality. By the same token, the Grand Chamber can be said to have failed to address the views of the Italian Constitutional Court in its most recent case law in this regard, as for example judgment no 181/2024, which introduced the concept of “constitutional tone”.

 

In the case, this view was supported by all the intervening Member States and the Commission with a more general argument aimed at underlying the difference between a general judicial review, which would involve the possibility, for the national judge, to question whether the designation of a country as ‘safe’ complies with the criteria laid down in Annex I of Directive 2013/32, and a more individualised assessment that the national judge will perform on whether the designation of a third country as a SCO does not apply to the individual(s) whose case is under examination, meaning that, for them and only for them, the country cannot be considered as ‘safe’. According to this argument, that in my opinion was best summarised at the hearing by the counsel for the German Government:

 

-          Directive 2013/32 distinguishes between the general designation of a country as a SCO, on the one hand, and the application of the concept in a specific case, on the other hand. The former assessment is based on general considerations, while the latter is based on the specific circumstances of the asylum seeker concerned.

-          Every national court has the power to perform the second assessment (application of the general designation of a country as a SCO to a specific case). However, the first assessment (general designation of a country as a SCO) must remain separate from this, and reserved to the Legislature. This is so because EU law leaves ‘leeway’ to Member States

-          If a national court believes that the general assessment that prompted the designation of a country as a SCO goes against the Constitution and/or EU law, it will have to raise a challenge of constitutionality before the constitutional court.

 

This was certainly taken into consideration by the AG in his opinion, which highlights the difference between a general challenge to the designation of a SCO and the decision, based on the specific circumstances of the case, that the designation does not apply to one or more individuals. This is probably the reason why, in his opinion, the AG never mentioned dis-application and rather chose to rely on the right to judicial protection (Article 47 of the Charter) and the right to an effective remedy (Article 46 of Directive 2013/32).

 

Against this backdrop, it seems as though the judgment does not really engage in these nuances. Indeed, the Grand Chamber refers to both dis-application and the right to an effective remedy when establishing that the national judge has the power to review the designation of a SCO. Yet, that part of the judgment may appear to be excessively straightforward. Indeed, the Court does not explicitly say that Article 37 and/or Annex I of Directive 2013/32 have direct effect, but merely implies this when referring to disapplication at § 63 of the judgment.

 

This omission is all the more apparent when considering that

 

-          the question as to whether Article 37 has direct effect was referred to the CJEU by the district Court of Bologna in case C-750/24 Ortega, but the Court suspended the examination of that request and gave priority to Alace and Canpelli;

-          in a different section of the judgment (at § 77) the Grand Chamber explicitly says that Articles 47 of the Charter and Article 46 of the directive have direct effect;

-          as I have already argued in my previous blog it does not seem too obvious that Article 37 qualifies as a provision of EU law capable of having direct effect on the basis of previous case law of the CJEU on the matter.

 

That said, it is also true that, leaving aside technicalities that may well be addressed in future cases, the answer of the Grand Chamber on the matter is clear: the need to ensure a full and fair examination of asylum claims at the European Union level requires that the national judge has the power to review each and every aspect of the request, including the designation of a country of origin as ‘safe’. In this sense, the judgment fits in the CJEU case law that has historically underscored the importance of the role of the national judge in the correct implementation of EU law.

 

The conclusion is the same as far as the answer to the fourth question is concerned, as it is indeed remarkable that the CJEU dismissed the “pragmatic” approach put forward by the Italian Government and resisted the “pressure” exerted by the intervening Member States and by the abrupt change of position of the Commission. The Grand Chamber also disavowed the opinion of the AG, which clearly favoured a loose interpretation of Article 37 and Annex I of Directive 2013/32, including on the basis of a questionable interpretation of the letter of the latter provision. Instead, the CJEU followed the defence, which had argued that ‘the letter of directive 2013/32, and especially its Annex I […] leaves no doubt as to the absolute impossibility to designate a country as a SCO whenever there are elements to hold that the country is, in fact, not safe for specific categories of people’. In so doing, it stressed the importance of the meaning of the hendiadys “generally” and “consistently”, saying that these terms require that the situation in a given country is “invariably” safe for its entire population.

 

Interestingly, while the reply to the first question can be read as a re-affirmation of the central role of the judiciary, the answer to the last question focusses on the prerogatives of the EU Legislature when exercising its discretion and striking a fair balance between the two conflicting objectives of Directive 2013/32; i.e. the interest to accelerate the examination of asylum claims vis-à-vis the duty to respect the basic rights of the asylum seekers.

 

For this reason, it seems to me that, rather than in terms of a conflict of powers (judiciary v executive) or institutions (EU v Member States), the judgment of 1 August 2025 should be read under the lenses of the tyranny of values. In this perspective, the issue is not as much which authority has the final say over an asylum claim, but whether, as a community based on the Rule of Law, the European Union and the Member States intend to place emphasis on pragmatic considerations regarding the burdensome effects of a large influx of asylum seekers, or whether the polar star is and remains the protection of fundamental rights - first and foremost the right to effective judicial protection.