Managing Migration the Italian Way II

Here we go again. Eight months after first exploring the legal contours of the “innovative” Italy-Albania Protocol and its temporary suspension following the Court of Rome’s refusal to validate the detention of the first group of asylum applicants transferred to Albania, this post turns to a new and potentially game-changing development: a preliminary reference to the European Court of Justice (ECJ), submitted by the Italian Court of Cassation on 20 June 2025 (and made available here). The referral raises doubts about the compatibility of the scheme with both the Return Directive (RD) and the Asylum Procedures Directive (APD). This move adds fresh legal uncertainty to a deal already under intense scrutiny and could significantly impact its implementation.

In what follows, I will first provide a brief overview of the changes made to the Protocol’s implementation following the initial judicial stalemate. I will then unpack the novelties of the most recent referral, and then explore what it might mean for the future of Italy’s experiment.

Revisiting the Italy-Albania deal

In the last post, I examined the early legal challenges to the Italy–Albania Protocol on offshore asylum processing, which enables Italy to conduct an accelerated border procedure in Albanian territory for individuals rescued in international waters, and allows for the return of rejected applicants. At the time, implementation of the Protocol had just begun: the facilities in Shengjin and Gjader were operational as of October 14, and two groups of asylum seekers had already been transferred to Albania.

The first judicial blow to the Protocol came on 18 October, when the Court of Rome refused to validate the detention of 12 asylum seekers from Egypt and Bangladesh. The Court found that the accelerated border procedure had been unlawfully applied, drawing on a recent ECJ judgment which prohibits designating a third country as a “safe country of origin” (SCoO) if territorial exceptions exist. Applying this reasoning analogically, the Court of Rome concluded that since Egypt and Bangladesh were listed with category-based exceptions, the applicants could not lawfully be subjected to the accelerated procedure. Consequently, their detention was deemed unlawful and the applicants were returned to Italy. In response, the government issued a law decree on 24 October amending the SCoO list by removing territorial exceptions, while retaining group-based ones. It also restructured the appeals process by rerouting judicial review of detention decisions through appellate courts, arguably to secure more favourable outcomes. This is where the previous analysis left off.

Since then, Italian courts have submitted a series of preliminary references to the ECJ, all essentially seeking clarification on whether the Court’s judgment of 4 October should also be interpreted as precluding category-based exceptions—not just territorial ones—in SCoO designations. The Advocate General’s Opinion, however, suggests that such exceptions may be permissible, provided certain conditions are met (for further analysis, see here).

While awaiting a final ruling, Italy has suspended the interception and transfer of asylum seekers to Albania. In the meantime, the nature of the facilities in Albania has shifted: to avoid leaving the centres empty, the government issued a new decree—later converted into law—authorising the transfer of irregular migrants from pre-removal detention centres on Italian territory (CPRs) to Albania, pending repatriation. This move has reignited legal debate, this time focusing on its compatibility with both the Return Directive and the Asylum Procedure Directive.

The legal core of the Cassation Court’s referral

The preliminary reference submitted by the Italian Court of Cassation raises two central legal questions regarding the compatibility of Italy’s implementation of the Protocol with EU law. First, it asks whether the RD precludes a national provision that permits the transfer of migrants from Italian CPRs to Albania—even in the absence of a concrete and identifiable prospect of return. Second, if such transfers are found compatible with EU law, the Court seeks clarification on whether Article 9(1) of APD, which requires that applicants be allowed to remain in the Member State for the purpose of the asylum procedure, prevents the detention of a migrant who, once transferred to Albania, submits an asylum request presumed to be merely instrumental in avoiding removal.

What stands out is that, unlike the previous referral, which took for granted the applicability of EU law to the asylum processing in Albania, the issue is now directly tied to the territorial reach of the EU acquis. Much of the EU’s secondary legislation on asylum and return is territorially bound, applying within the territory and the geographical borders of Member States. Yet under the Protocol, migrants are transferred to a third country, Albania, even though the agreement formally commits to applying Italian and EU standards “as if” the individuals were on Italian soil. For the first time, the Court of Cassation’s question acknowledges a key legal tension: does the formal extension of EU norms to a third country, through a legal fiction that equates its territory with that of a Member State via a bilateral agreement, suffice to guarantee compliance with EU law?

An insufficient legal fiction

The Italian court does not appear to agree. Much in line with arguments made elsewhere, the Italian judges acknowledged that there exists a sufficiently close connection between the EU legal order and the situation of migrants in Albania—whether they are applicants for international protection or third-country nationals awaiting removal. This connection entails that the standards laid down in EU directives, even when unilaterally applied in a third country, must be concretely respected. Failure to do so risks undermining the EU’s objective of establishing a Common European Asylum System (Article 78 TFEU) and a unified immigration policy, including the removal and repatriation of unauthorized residents (Article 79 TFEU), which appears not to be fully respected in the procedures currently carried out in Albania.

Regarding compliance with the standards set out in the RD, the Court observed that it remains unclear how the transfer and detention of returnees in Albania serves the Directive’s objective of facilitating repatriation to their countries of origin. This uncertainty raises concerns about the legality of such deprivation of liberty. With regard to the APD, which becomes applicable if returnees transferred from Italian CPRs to Albania submit an application for international protection there, the Court expressed doubts about whether the legal fiction of treating the centres in Albania as Italian border zones is sufficient to guarantee the applicant’s right to remain in the territory—including at the borders—while their claim is being processed. These concerns are particularly acute given the likely lower level of safeguards and rights available in a third country, despite the formal territorial equivalence established by the agreement (for an analysis of the main issues, see here).

In sum, the referral by the Italian Court of Cassation brings to the forefront the unresolved legal complexities surrounding the extraterritorial application of EU asylum and migration law. At its core, the case tests the limits of the EU legal order’s territorial scope and questions whether legal fictions embedded in bilateral agreements can substitute for genuine guarantees and ensure effective access to rights derived from EU law. By scrutinizing the Albania Protocol in light of both the RD and the APD’s scope of application, the Court signals its concern that the mere formal invocation of EU standards is insufficient when there is a tangible risk of diminished rights and safeguards. Ultimately, the referral challenges the assumption that procedural compliance on paper can compensate for substantive shortcomings in practice, particularly when legal obligations are displaced beyond EU borders.

The future of externalized return hubs  

On 11 March 2025, the European Commission presented a proposal to repeal the Return Directive and replace it with a Regulation establishing a common system for the return of third-country nationals staying illegally in the Union. The proposal broadens the definition of “country of return,” allowing Member States to transfer returnees to third countries they have an agreement or arrangement with to accept the third-country national (Article 4(3)(g)). If adopted in its current form, this provision could pave the way for the establishment of so-called return hubs—a controversial policy tool that has already drawn comparisons to the Italy–Albania agreement. However, as the Court of Cassation itself acknowledged, substantive differences remain between the two models, meaning that the ECJ’s answer to the current referral may offer only limited guidance for addressing the broader legal issues raised by return hubs as envisioned under the proposed Regulation.

The concept behind return hubs is to outsource the return process to a third country that is not the returnee’s country of origin, but which may either facilitate their onward journey or permit them to remain. This would be typically achieved through the issuance of a return decision directed at that third country, meaning that once the individual is transferred, the Member State’s legal obligations would, in principle, cease. By contrast, as discussed, the Italy–Albania model involves the transfer of migrants to Albania pending repatriation to their country of origin, while Italy retains jurisdiction and, consequently, responsibility. Notably, no return decision is issued for Albania; instead, returnees detained in Italian CPRs are treated as if they were internally transferred to another facility within national territory. This distinction implies that even if the ECJ were to find the Italian scheme incompatible with EU law, alternative models for shifting responsibility to third countries—whether in the context of asylum or return—would remain open for debate, despite their high cost and predictable low long-term effectiveness.

An entrenched commitment to shift responsibility

What this entire legal and political saga reveals is that the EU’s preoccupation with returns and externalisation is not a transient trend but a deeply entrenched policy orientation. Despite judicial pushback and the legal uncertainties exposed by cases such as the UK-Rwanda and the Italy–Albania agreement, Member States continue to pursue increasingly complex mechanisms to shift responsibility beyond EU borders. This reflects a persistent effort to reconcile restrictive migration control objectives with the legal constraints imposed by (EU) law. Ultimately, whether the ECJ’s forthcoming judgment will meaningfully shape or constrain these strategies remains to be seen. What is certain, however, is that the overarching ambition to prevent access to EU territory—and thereby to effective protection—continues to outweigh efforts to develop credible, humane, and sustainable approaches to welcoming people within the Union.