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.
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.
.jpg){kind=link}
No comments:
Post a Comment