Tuesday, 31 March 2026

Detention by Delay? A Structural Paradox in EU Migration Law

 


 

Miguel Alconero Bravo, Predoctoral Research Fellow (Formación de Personal Investigador – FPI) at the University of Valladolid*

*Part of the Research Project “Proceso Penal y Espacio de Libertad, Seguridad y Justicia: Garantías, Cooperación Transfronteriza y Digitalización” (Ref. PID2023 – 152074NB – I00). 

Photo credit: berthgmn via wikimedia commons

 

Introduction

 

Predicating the detention of a migrant on circumstances entirely beyond his or her control seems prima facie unjust. Indeed, even a cursory reading of the provisions of EU law empowering Member States to extend the detention of a foreign national due to delays in obtaining the necessary documentation from third countries unmistakably exposes the inherent disproportionality of such a measure.

 

Ordering detention v. Extending detention

 

It must first be emphasised that both Article 15(6) of the Return Directive and Article 32(3) of the Commission proposal to overhaul EU return procedures enable Member States solely to extend a period of deprivation of liberty that has already been lawfully ordered.

 

Accordingly, where the conditions required under EU law for the detention of a third-country national are not satisfied in a given factual scenario, it is legally immaterial whether any of the grounds permitting a State to prolong that measure are present. If a person could not lawfully have been deprived of liberty from the outset, there is, fundamentally, no valid detention susceptible of being extended.

 

In this regard, Article 15(5) of the Return Directive states that “detention shall be maintained for as long a period as the conditions laid down in paragraph 1 are fulfilled and it is necessary to ensure successful removal”, whereas Article 32(1) of the proposal for a Return Regulation provides that “detention shall be maintained for as short a period as possible and for as long as the conditions laid down in Article 29 are fulfilled […]”.

 

Reference must therefore be made to Article 15(1) of the Return Directive, which adopts a numerus apertus approach. This provision refers in particular to two situations that may lead to the detention of a foreign national pending removal: the existence of a risk of absconding, and the possibility of identifying conduct through which the migrant seeks to avoid or hamper the preparation of return or the removal process.

 

Similarly, Article 29(3) of the proposed Regulation sets out, in exhaustive terms, the only situations in which the requesting State may order the detention of an irregularly staying third-country national: (a) where there is a risk of absconding determined in accordance with Article 30; (b) where the third-country national avoids or hampers the preparation of the return or the removal process; (c) where the third-country national poses security risks in accordance with Article 16; (d) in order to determine or verify his or her identity or nationality; (e) in the event of non-compliance with the measures ordered pursuant to Article 31.

 

Thus, the vast majority of the grounds for detention laid down in the Return Directive and in the proposed Regulation relate to scenarios in which the third-country national concerned has sought to frustrate the enforcement of return or the removal process. In essence, deprivation of liberty under both instruments is generally linked to some form of non-cooperation by the person concerned.

 

In this context, it must be stressed that construing the foreign national's lack of cooperation as a ground for extending detention – as Article 15(6)(a) of the Return Directive and Article 32(3) of the proposed Regulation do – makes it easier to postulate that the ground originally justifying detention still subsists. In other words, such a construction allows the authorities to establish the presence of almost any of the grounds set out in Article 15(1) of the return Directive and in Article 29(3) of the proposed Regulation for the purpose of ordering the detention of a migrant.

 

As a matter of fact, when a State invokes the “lack of cooperation by the third-country national concerned” in order to prolong detention, it is essentially relying on a concept whose indeterminacy allows national authorities to subsume within it many of the very grounds on which they could initially have relied to deprive that person of liberty.

 

In sum, extending the detention of a foreign national as a consequence of his or her lack of cooperation will, in most cases, make it possible to ascertain that the grounds for detention set out in EU law continue to be met.

 

Nonetheless, an exception must be made in respect of Articles 29(3)(c) and 29(3)(d) of the proposed Regulation –namely, where the person concerned is a third-country national deemed to pose a security risk, and where detention is ordered for the purpose of determining or verifying the foreign national’s identity or nationality.

 

In that respect, it should be noted that detention ordered on the grounds that a migrant poses a risk to public order, public security or national security bears, strictly speaking, no relation to the degree of cooperation he or she may have been willing to display during the return procedure. Irrespective of the individual’s willingness to cooperate, once a foreign national has been deprived of liberty on the basis that he or she is regarded as a threat to a State’s public order, public security or national security, detention will persist for as long as that classification remains in force.

 

That being so, authorities cannot invoke a migrant’s lack of cooperation to extend detention where the initial deprivation of liberty was based on security grounds or on the need to verify identity, as those circumstances do not, as such, relate to the individual’s conduct during the return procedure (unless a genuine and subsequent lack of cooperation actually materialises). Nevertheless, it is equally true that, in our view, Brussels errs in treating said situations as grounds for detention in the first place.

 

Against this backdrop, the pivotal question is as follows: under which of the grounds set out in either Article 15(1) of the Return Directive or Article 29(3) of the proposed Regulation could a Member State’s intention to prolong the detention of a third-country national be classified when that intention is based, purely and simply, on the likelihood that the procedure will take longer owing to delays in obtaining the required documentation from third countries?

 

As far as the Return Directive is concerned, the answer may lie in the numerus apertus structure of Article 15(1). Article 29(3) of the proposed Regulation, however, does not, under any circumstances, permit a third-country national to be deprived of liberty solely on the basis of delays in obtaining documents from third countries.

 

The conclusion follows inexorably. The wording of Article 32(3) of the proposed Regulation allowing deprivation of liberty to be prolonged “where the return procedure is likely to last longer owing to […] delays in obtaining the necessary documentation from third countries” ought to be removed. If EU law does not recognise the possibility of detaining a foreign national on that ground, it is a fortiori manifestly contrary to the principle of proportionality that such delays should then be relied upon to justify prolonging the duration of a custodial measure.

 

A structural contradiction

 

It must be acknowledged that both Article 32(3) of the proposed Return Regulation and Article 15(6)(b) of the Return Directive pursue an entirely legitimate aim: namely, to prevent, through the use of all available means, a situation in which enforcement of the return decision becomes impossible as a result of a temporary delay in receiving the necessary documentation from the requested State.

 

That said, it is crucial to recall that, in order for the detention of a third-country national to be prolonged, there must first exist a custodial measure resting on a valid legal basis. The analysis must therefore begin by determining whether in the factual circumstances of the case the substantive conditions laid down in EU law for depriving a migrant of liberty are actually satisfied.

 

Against this background, neither Article 15(1) of the Return Directive nor Article 29(3) of the proposed Regulation treats a mere delay in the completion of the administrative or diplomatic formalities necessary to obtain the required documentation from third countries as a circumstance capable of justifying an initial deprivation of liberty.

 

This gives rise to a stark legal contradiction: if EU law precludes the initial detention of a third-country national on the basis of administrative delays in obtaining the necessary documentation from third countries, why does Brussels permit its Member States to extend detention on that very ground? Although such delays cannot justify the initial deprivation of liberty, they may nonetheless serve to prolong the detention of a migrant.

 

The result is a structural paradox whereby a custodial measure that should come to an end from the moment the substantive conditions underpinning detention are no longer met is nevertheless lawfully prolonged by the requesting State under the provisions governing extension.

 

That tension becomes all the more striking when viewed in the light of other instruments of EU migration law. In particular, Article 11(1) of Directive 2024/1346 provides that “delays in administrative procedures that cannot be attributed to the applicant shall not justify a continuation of detention”.

 

For all these reasons, if the EU wished to preserve this ground for prolonging detention in the proposed Regulation, its express incorporation into Article 29(3)  as a ground, not just for extending, but also for ordering detention would be necessary.

 

As regards Directive 2008/115, the only interpretative solution would be to conclude that, by merely identifying the risk of absconding and the possibility that the individual may avoid or hamper the preparation of return as the two grounds that, in particular, entitle a State to deprive a migrant of liberty, that instrument implicitly leaves room for the authorities to rely on other situations as independent grounds for detention. One such additional scenario would be, precisely, delays in obtaining the necessary documentation from third countries.

 

However, both of those alternatives must be categorically rejected. Otherwise, the ultima ratio nature traditionally associated with the detention of an irregularly staying third-country national would be severely compromised.

 

In any event, the release of the individual concerned does not prevent the authorities from resorting to less restrictive measures designed to secure the enforcement of return pending receipt of the necessary documentation.

 

In this regard, Article 31 of the proposed Regulation explicitly provides for various alternatives to detention, which must be “proportionate to the level of the risk of absconding assessed in accordance with Article 30”. Furthermore, Article 23 of said proposal provides for a series of options which, by restricting the individual’s geographical mobility, are capable of ensuring his or her availability during the removal process.

 

Mahdi: When can detention be extended?

 

The Court of Justice of the European Union (CJEU) clarified in Mahdi the scope of the assessment that national courts must undertake before authorising any extension of a detention period on the basis of Article 15(6) of the Return Directive

 

In order to rule on the questions referred by the Administrativen sad Sofia-grad, the CJEU began by identifying the defining features of Article 15 of the Return Directive. It emphasised, in particular, that this provision is both “unconditional” and “sufficiently precise”, qualities which explain why its implementation by the Member States requires “no other specific elements”.

 

The Court then turned to the provisions whose interpretation had been requested by the referring court, namely Articles 15(3) and 15(6) of the Return Directive.

 

Concerning Article 15(3), the Court underlined the clarity with which its wording indicates that any detention measure prolonged over time must be subject to the “supervision” of a judicial authority. Yet the CJEU also acknowledged that the provision does not specify the precise nature of that examination, thereby making it necessary to “recall the rules deriving from Article 15 […]” in order to complete the analysis required by the question raised by the Bulgarian court.

 

It was at that stage that the Court brought Article 15(6) of Directive 2008/115 into the discussion.

 

Significantly, however, the CJEU did not confine itself to identifying the consequences flowing from the substantive conditions set out in that provision. Instead, it examined Article 15(6) in the light of Article 15(4) of the same Directive.

 

In so doing, the Court of Justice of the European Union introduced two decisive elements into the judicial review of the deprivation of liberty of an irregularly staying third-country national.

 

First, it held that, “at the time of the national court’s review of the lawfulness of

detention”, there must be “a real prospect that the removal can be carried out successfully”, having regard to the time limits laid down in Articles 15(5) and 15(6) of Directive 2008/115.

 

Secondly, the Court established that the authority determining whether the individual’s detention should be extended or whether he or she should be released must re-examine the substantive conditions laid down in Article 15(1) of the Return Directive. Crucially, this entails verifying that the grounds “which have formed the basis for the initial decision to detain the third-country national concerned” continue to subsist.

 

This point is of central importance to the present discussion, since Article 15(1) of the Return Directive makes no reference whatsoever to the possibility of detaining a foreign national solely on account of delays by third countries in issuing the necessary documentation. It follows that reliance on such a ground does not relieve the competent authorities of the burden of demonstrating, first, that there remains a “reasonable prospect of removal” and, secondly, that the substantive grounds which justified the initial deprivation of liberty continue to subsist.

 

The disappearance of either the reasonable prospect of removal or the substantive grounds that initially justified the detention of the third-country national necessarily renders that deprivation of liberty unlawful, as Article 15(4) of the Return Directive makes clear. That conclusion is not altered by the mere fact that the authorities continue to await the necessary documentation from other States.

 

Monday, 30 March 2026

Clash between gender data vs Hungarian personal data register: Can an existing Hungarian registry system prevent the enforcement of trans rights based on the GDPR?


 

Attila Szabó, PhD

Head of Legal Aid Service, Hungarian Civil Liberties Union

 

*The author assisted the lawyer representing the person concerned as an advisor in the Hungarian case analyzed in the text.  The author also used artificial intelligence to prepare the English version of the text.

 

Photo credit: Jorge Franganillo, via Wikimedia commons

 

In recent years, the Court of Justice of the European Union has increasingly engaged with trans rights. European constitutional community has followed this development. 

It appears that despite the consolidation of a GDPR interpretation aligned with trans rights, and thus with human dignity, Hungarian courts fail to understand that gender identity is not only a matter of self-determination, but also one of data accuracy. If someone presents and lives as a woman, then from the perspective of data accuracy she must also be treated as a woman, since this is the accurate data. The highest Hungarian court, however, sees this differently, and with its decision on the matter, it violates EU law.

Between 2024 and 2025, the Court of Justice of the European Union (CJEU) reshaped the landscape of trans rights in the EU through a remarkable line of cases: Mirin, Mousse, Deldits and Shipov. Taken together, these rulings reveal a structural shift in the CJEU’s approach. The Court increasingly speaks the language of “gender identity” rather than “gender reassignment,” signalling a move away from medicalised understandings of trans status and opening space for non-binary recognition. It integrates ECtHR standards as a constitutional floor while embedding trans rights across multiple doctrinal pillars: EU citizenship, free movement, privacy, equality, and data protection. What emerges is not a single breakthrough, but a coherent jurisprudential arc. One that justifies speaking of a significant doctrinal shift of trans rights in EU law.

Summary of the Hungarian Kúria’s Decision 

The claimant requested the rectification of the “sex” entry in the Hungarian personal data and address register from “male” (as recorded in the civil registry at birth) to “female,” relying primarily on Article 16 GDPR (right to rectification) and explicitly invoking the CJEU’s judgment in Deldits. The claimant argued that the register in question records “sex”, not “sex at birth,” and therefore should reflect lived (social) gender identity. Since her appearance and social relations objectively correspond to a female identity, the currently recorded data were inaccurate within the meaning of Article 5 (1) d) and Article 16 of GDPR. She maintained that the data protection authority should assess whether the recorded data correspond to reality as experienced and perceived, and that EU law requires rectification even if domestic civil registry law remains unchanged.

Both the first-instance court and, on review, the Kúria (Hungarian Supreme Court) rejected the claim. (The decisions have not yet been made public; this blog post provides the first summary of them.) The decisive reasoning was structural: the personal data and address register is a secondary, derivative register whose “sex” entry is based directly on the birth registry. Under Hungarian law, the birth registry records “sex at birth,” defined biologically. Since the personal data register derives this data from the civil registry, it cannot diverge from it without undermining legal certainty and the authenticity of public registers. The Kúria held that the register does not record “lived gender identity” at all; therefore, the data cannot be considered inaccurate merely because the claimant’s current gender identity differs from the birth record. In its reading of Deldits, the GDPR right to rectification applies only to data that are inaccurate within the meaning and function of the specific register concerned. Article 16 GDPR cannot be interpreted as obliging an authority to insert new categories of data (e.g. lived gender identity) without explicit statutory authorisation, nor to assign a different substantive meaning to an existing category (“sex at birth”). The Court therefore concluded that no inaccuracy existed and that rectification was not required.

Mistakes in the decision

The conceptual distinction between “sex at birth” and “sex” undermines the exclusivity claim. The birth registry records a historical biological fact at the time of birth. By contrast, the personal data and address register functions as an operational identification database used for everyday legal and administrative interactions. Accuracy in this context serves identification and legal certainty in present-day relations. If the data recorded there do not reflect the individual’s lived and socially recognised gender, they may fail the accuracy requirement precisely because they hinder reliable identification. The fact that one dataset originates historically from another does not transform the original entry into an immutable legal truth for all future processing contexts.

Secondly, Hungarian law itself does not establish that the civil registry is the sole permissible source of “sex” data in all registers. If it would have been so it would be absolutely unchangeable. However it is not, since the statutory framework governing the personal data and address register allows updates based on legally valid rectification requests. It is fully in line with GDPR. Moreover, even Hungarian constitutional jurisprudence has recognised that legal acknowledgment of gender identity, at least outside the civil registry context, may be compatible with the Hungarian Fundamental Law. This demonstrates that the legal system does not treat the birth entry as metaphysically definitive, but as one administrative record among others.

Finally, the “hierarchy of registers” argument reverses the logic of legal certainty. Legal certainty is not preserved by maintaining inter-database consistency at the price of factual inaccuracy. Rather, certainty requires that state records correspond to verifiable social and legal reality. If necessary, consistency between registers can be achieved by differentiating between “sex at birth” (retained in the birth registry) and current “sex” or gender identity (reflected in identification databases). EU law does not require uniformity of terminology across all databases; it requires accuracy, proportionality, and effective protection of fundamental rights. Therefore, the proposition that only the birth registry may serve as the lawful source of sex-related data is neither compelled by domestic law nor compatible with the GDPR as interpreted by the Court of Justice.

The Decision in the Context of European Law

This ruling stands in notable tension with the emerging CJEU jurisprudence represented by the Deldits, Mirin, Mousse and Shipov cases. In Deldits, the CJEU held that where a register contains personal data relating to gender identity, that data must be rectified if inaccurate, and that Member States may not impose disproportionate evidentiary burdens (such as proof of surgery). The Hungarian Kúria distinguished Deldits on the basis that the asylum register there functioned as a primary identity register, whereas the Hungarian personal data register merely mirrors the civil registry’s birth-sex entry. The core of Kúria's reasoning is therefore ontological: if a register is designed to record biological sex at birth, then a divergence from lived gender identity does not render it “inaccurate.”

This is, of course, a misconception: the personal data register, which is distinct from the civil (birth) registry, exists precisely to record the data necessary for identification. In most cases, those data derive from the civil registry; however, in the case of trans persons, they do not necessarily follow from the birth registry but from factual circumstances that, through a rectification procedure, could also become officially documented facts. This state database can record birth sex data and accurate, actual data too, in parallel. Kúria argues that the registry only processes data relating to "sex" and cannot process data relating to birth gender and current gender without a change in legislation. This is true, of course, but the data controller can, in such cases, process only data relating to current gender in the registry that exists alongside the birth registry system.

Nor is Kúria's argument persuasive that it cannot order the processing of “new” data. The personal data register can, within the existing legal framework, be modified technically and administratively if in its current form it does not comply with the requirements flowing from the GDPR and EU law. The Kúria therefore did not give effect to EU law, but rather to domestic practical constraints: it effectively treated a functional system as if it were a legal norm, even though in reality such a system should adapt to legal norms, not prevent their enforcement.

From an EU law perspective, however, this formalistic register-based distinction raises deeper questions. The recent CJEU trend has emphasised substance over classification: Mirin prioritised the practical effectiveness of EU citizenship and identity coherence across registers; Mousse treated gender-related data as protected personal data subject to strict necessity and proportionality review; and Deldits framed gender identity as a legally relevant dimension of accuracy under the GDPR. Against this background, the Hungarian decision represents a restrictive reading of Article 16 GDPR, confining rectification to internal consistency within a nationally defined registry hierarchy.

In this case, we will still turn to the Hungarian Constitutional Court. However, if that body also fails to restore the possibility of effective legal enforcement under the Hungarian Fundamental Law and the binding EU law applicable on that basis, then, besides applying to the European Court of Human Rights, the only remaining option will be for the European Commission to initiate infringement proceedings and thereby compel Hungary to comply with the binding requirements of the GDPR and the Charter of Fundamental Rights. And yes, if necessary, let them add another field to the personal data register system.

Following the very recent Shipov decision, the situation is even clearer: the position of the Hungarian Supreme Court is completely untenable. In that case, the court ruled that it violates the right to free movement if a person is unable to identify themselves in another Member State with an identity document corresponding to their true gender. The current Hungarian case highlights that this violates not only the right to free movement but also the GDPR’s data accuracy rules. After all, inaccuracy is not only a problem when someone travels to another Member State. The two cases are thus based on different legal arguments, but they point to the same thing: inaccuracy causes privacy difficulties that violate the right to private life protected by the Charter of Fundamental Rights.

 

 

 

Sunday, 29 March 2026

The Primacy of EU Law in Bulgaria after CJEU’s Judgment in Case C‑56/25: A Thorny Path Ahead

 


Dr Radosveta Vassileva, Adjunct Senior Research Fellow, UCD Sutherland School of Law

Photo credit: Raggatt2000, via wikimedia commons

On 12 February 2026, the CJEU delivered its judgment in Case C-56/25 which concerns a preliminary reference from Bulgaria raising a question about the application of the principle of primacy of EU law in view of a provision of the Procedural Rules of Bulgaria’s Constitutional Court. The CJEU judgment was rendered without an Opinion by the Advocate General, indicating that, in the eyes of the CJEU, the question was neither new, nor difficult. Moreover, the CJEU quotes its settled case law on the primacy of EU law.

However, a closer look at the Bulgarian legislation, which gave rise to the question by the Bulgarian court, especially against the broader country-specific context, reveals that the judgment seems to foreshadow inevitable conflicts between the Bulgarian and the EU legal orders in the future. From a Bulgarian perspective, the place of EU law in the hierarchy of norms is far from being conclusively determined. In parallel, the judgment demonstrates some of the flaws of a controversial constitutional reform carried out in Bulgaria in 2023, whose grand innovation was the ‘individual constitutional complaint’, which led to the hierarchy of norms dilemma of the referring court.

‘The chicken-or-the egg’ problem of the Bulgarian court

The preliminary reference was made in 2025 by the Sofia City Court, acting as a first instance in criminal proceedings, which seemingly faced a ‘chicken-or-egg’ dilemma.

The Sofia City Court was concerned that a provision of national law relevant to the qualification of the criminal offence and, respectively, the punishment in the criminal proceedings before it violated both Bulgaria’s Constitution and EU law – namely, Article 4 of Council Framework Decision 2004/757/JHA of 25 October 2004 laying down minimum provisions on the constituent elements of criminal acts and penalties in the field of illicit drug trafficking.

The Sofia City Court deemed that it was more appropriate to ask the Constitutional Court to rule on the constitutionality of the contested provision of national criminal law before it made a preliminary reference to the CJEU about the compatibility of Bulgarian criminal law with the aforementioned council framework decision. However, in its eyes, there was a catch – Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court requires that a request to it

…must contain a reasoned assessment of the applicable law, including of the consequences of the application of EU law where the contested provision or act comes within its scope.

In this context, the Sofia City Court asked the CJEU the following:

Are Article 267 TFEU, Article 94(b) of the Rules of Procedure of the Court … and the principle of the primacy of EU law … to be interpreted as meaning that, where a national court has doubts as to the compatibility of a provision of national law with EU law, and is at the same time convinced that that provision of law is [contrary to the national Constitution], it is obliged or entitled, before submitting its request for a preliminary ruling, to establish whether that provision of national law is indeed applicable in the main proceedings by making an application to the… Constitutional Court… for a declaration as to its unconstitutionality?(para 27 of judgment).

Contextual background: ‘the individual constitutional complaint’ as the grand innovation of the 2023 constitutional reform

The significance of the question by the Sofia City Court can be better appreciated against the broader context of the Bulgarian constitutional reform of 2023, which, regrettably, is neither explained in the text of the preliminary reference itself, nor in the CJEU judgment. This reform, however, made Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court, which is at the heart of the hierarchy of norms dilemma of the Sofia City Court, a legal irritant from an EU law perspective.

Before the 2023 reform, the opportunities to request a review by the Constitutional Court were ‘significantly limited and depend[ed] on the political climate in the country’ (see Radosveta Vassileva, ‘A Perfect Storm: The Extraordinary Constitutional Attack against the Istanbul Convention in Bulgaria’ (2022) 1 Osteuropa Recht 84). According to the previous wording of Article 150(1) of the Bulgarian Constitution, the Constitutional Court could only be approached by one-fifth of the Members of Parliament, the President, the Council of Ministers, the Supreme Court of Cassation, the Supreme Administrative Court, or the General Prosecutor. In some limited cases, it could be approached by the Ombudsman or the Supreme Bar Council (see the previous wording of Articles 150(3) and 150(4) of the Constitution).

To this end, one of the grand ‘innovations’ of the 2023 constitutional reform was the introduction of the ‘individual constitutional complaint’. The name given by the Bulgarian legislator to this new procedure, nevertheless, seems to be a misnomer because a private citizen can neither directly ask the Constitutional Court for constitutional review nor appeal court decisions there on constitutional grounds. The new mechanism for constitutional review, in essence, mimics the preliminary reference procedure before the CJEU. In fact, as argued in the explanatory memorandum accompanying the reform bill of 2023, ‘the situation in Bulgaria [was] paradoxical because any court (judge) could [make a preliminary reference to the CJEU], but could not turn directly to the national constitutional jurisdiction’ (p 17 of the aforementioned explanatory memorandum).

Following the 2023 reform, Article 150(2) of the Constitution stipulates:

Any court, at the request of a party to the case or on its own initiative, may refer to the Constitutional Court a request to establish an incompatibility between a law applicable to the specific case and the Constitution. The proceedings in the case shall continue, and the court, whose decision is final, hands down its judgment after the proceedings pending before the Constitutional Court have been concluded.

It is important to note that the usage of ‘may’ in the wording implies that the judge or the judicial panel in question has discretion over whether to honour such request by a party in the proceedings. In practice, it is common for courts to choose not to honour requests for preliminary references to the CJEU and/or requests to approach the Constitutional Court under the new wording of Article 150(2) coming from the parties. In other words, there are cases in which judges turn a blind eye to the likely incompatibility between Bulgarian legislation, on the one hand, and EU law and the Constitution, on the other. There are also cases in which judges can be seen in a more activist role, making references when the parties concerned did not argue any such incompatibility themselves.

The ‘ticking procedural time bomb’ left by the Bulgarian legislator

Bulgaria’s Constitution, adopted in 1991, clearly specifies the hierarchy of norms. Its Article 5(1) explicitly states that it is the ‘supreme law’, while its Article 5(3) stipulates:

International treaties, ratified in accordance with the constitutional procedure, promulgated, and having entered into force for the Republic of Bulgaria, are part of the country’s domestic law. They take precedence over those provisions of domestic legislation that contradict them.

Hence, the Constitution takes precedence over any international treaty and Bulgarian law while international treaties take precedence over Bulgarian laws that contradict them. Bulgaria has, of course, ratified its EU Accession Treaty of 2005.

Furthermore, the status and place of CJEU case law in the hierarchy of norms is subject to debate from a national perspective. CJEU’s case law is not defined as binding in Bulgarian legislation, unlike the case law of the Constitutional Court or the decrees and decisions on interpretation by the country’s supreme courts. However, the Code of Civil Procedure allows cassation on the grounds of violation of the case law of the CJEU (see Article 280(1), point 2).

It is notable that throughout the years, Bulgaria’s Constitutional Court has tried to avoid direct confrontation with the CJEU. In its own case law, however, it has argued that it is up to each entity requesting constitutional review to determine the applicable law, including EU law, to the proceedings before it. It has also stressed that the determination of the applicable law should always precede requests for constitutional review (see, for instance, the detailed reasoning in Ruling 2 of 24 February 2022 by the Constitutional Court).

In this light, the ‘individual constitutional complaint’ can be seen as a ‘ticking procedural time bomb’ since the reform of 2023 left Article 5 of the Constitution untouched, while expanding the jurisdiction of the Constitutional Court itself, which can now receive requests directly from judges. In practice, this ‘individual complaint’ means a chance for more frequent encounters and, respectively, potential clashes between Bulgarian constitutional law and EU law in ordinary Bulgarian courts.

The legislative choice not to tamper with Article 5, in turn, was informed by the restrictions on amendments imposed by the Constitution itself – any amendments to Article 5 require the convocation of a ‘grand national assembly’ as opposed to a ‘regular national assembly’ (on the difference between them, see here; see also Article 158 of the Constitution). Grand national assemblies are notoriously difficult to convene. The Constitution of 1991 was adopted by such an assembly.

CJEU’s judgment as an exercise in judicial diplomacy

In the judgment in Case C-56/25, the CJEU held that Article 267 TFEU, the principle of the primacy of EU law and Article 94(b) of the Rules of Procedure of the Court of Justice do not preclude procedural rules, such as Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court (para 61 of judgment). Leading authorities on the rule of law like Professor Laurent Pech have already called the judgment ‘unusually generous’, especially in view of challenges to CJEU’s jurisdiction from other EU member states, such as Poland.

Yet, when one reads the reasoning of the judgment in light of the particularities of the Bulgarian constitutional order, it seems that the ‘generosity’ came with a few price tags – namely, demands for a voluntary surrender of jurisdiction by the Bulgarian Constitutional Court. These requests, nevertheless, may not bear the fruit intended by the CJEU, precisely because they seem oblivious of context, including national legislation.

One of the reasons why the CJEU was not troubled by the requirements of Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court was the Constitutional Court’s alleged obligation to make preliminary references itself. Namely, the CJEU stated: ‘…if a constitutional court is seised of a request for a review of the constitutionality of a provision of national law coming within the scope of EU law, that court is in principle obliged to make a reference to the Court of Justice for a preliminary ruling, in accordance with the third paragraph of Article 267 TFEU…’ (para 57 of judgment). Bulgaria’s Constitutional Court, however, has no record of making preliminary references to the CJEU and it is highly unlikely that it will endeavour to make such references in the future if an amendment to Article 5 of the Constitution is not made. That is because the Constitutional Court sees its own role solely as the guardian of the Bulgarian Constitution. First, the Constitutional Court has the monopoly on authoritative, binding interpretation of the Constitution (Article 149(1), point 1 of the Constitution). Second, in case of conflict between the Constitution and EU law, the Constitution prevails according to its own text. Third, as highlighted above, the Constitutional Court has stated in its own settled case law that the establishment of the applicable EU law is in the prerogative of the court examining the dispute on the merits. On the one hand, this approach protects the Constitutional Court’s jurisdiction. On the other hand, it prevents direct confrontation with the CJEU – it is up to the ordinary courts to identify and raise any questions about the compatibility between EU law and constitutional law and to consider how such conflict affects the outcome of the case.

It is also interesting that the CJEU reminds that a ‘national court is required, in order to ensure the full effectiveness of the rules of EU law, to disregard, in the dispute before it, the rulings of a national constitutional court which refuses to give effect to a judgment given by way of a preliminary ruling by the Court of Justice…’ (para 59 of judgment). From an EU perspective, this approach has numerous merits, such as empowering judges in lower courts to disregard controversial decisions by constitutional courts, especially when they are unlawfully composed. Nevertheless, in the Bulgarian context, beyond the invitation to disregard key express provisions of the Bulgarian Constitution, this conclusion ignores important practical aspects. For example, pursuant to Article 280(1), point 2 of the Bulgarian Code of Civil Procedure, non-compliance with case law by the Constitutional Court serves as grounds for cassation. While the Code of Criminal Procedure does not contain such explicit reference to constitutional case law, its Article 348 allows cassation for ‘violations of the law’ broadly conceived.

A thorny path ahead for the primacy of EU law

Overall, the path ahead for the primacy of EU law in Bulgaria appears rather thorny. CJEU’s judgment in Case C-56/25 may be seen as a precursor to a series of legal and political challenges. It foreshadows inevitable clashes between EU law and Bulgarian constitutional law in ordinary courts, which are facilitated by the shortcomings of the 2023 constitutional reform.

It is also doubtful to what extent the approach, which the CJEU endorses in the judgment, will contribute to affirming the primacy of EU law in Bulgaria. Ensuring such primacy on paper requires important constitutional amendments by a grand national assembly as well as a large-scale legislative reform. It also necessitates a change of mentalité in practice on behalf of both Bulgarian judges and the CJEU itself.

First, CJEU case law remains terra incognita for many jurists because of their educational and professional background. Moreover, for more than a decade following Bulgaria’s EU accession in 2007, preliminary references from the country came from a very small circle of judges (see Aleksander Kornezov, ‘Ten years of preliminary references – a critical review and appraisal’(2017) Evropeiski praven pregled).

Second, between 2018-2022, a vast number of preliminary references were made by Bulgaria’s specialised criminal court, which had all features of an extraordinary repressive tribunal, and used this EU law mechanism primarily as a tool to affirm its legitimacy in light of criticism of its abusive practices. In 2022, this court was dissolved for good for undermining the rule of law upon the proposal by a short-lived opposition government and after years of public discontent. As a result, doubts about the status of its case law and the EU judgments resulting from its references, which are intertwined with it, are on the rise. So is the bitter feeling that the CJEU empowered ‘non-judges’ the way it did vis-à-vis Poland (see here).

Third, the CJEU has acted as a Pontius Pilate, washing its hands of responsibility, on key matters pertaining to Bulgaria’s rule of law backsliding and human rights abuses. It has a record of prioritising formalism and refusing to address the issues at their core, thus discouraging judges from raising politically sensitive questions (on judicial independence, see here; on standards of proof in pretrial detention, see here).

To what use is then primacy for primacy’s sake? This is a question which many judges from Bulgaria’s ordinary courts may ask in the aftermath of CJEU’s judgment in Case C-56/25.

 

The author would like to thank Prof Laurent Pech for his helpful comments and suggestions on an earlier draft.

Friday, 20 March 2026

The Chios Incident: Echoes of Pylos Humanitarian Disaster and Greece's Criminalization of Solidarity


 

Georgios Athanasiou, PhD Researcher, University of Antwerp

Photo credit: Julian Lupyan, via Wikimedia Commons

The Chios migrant boat shipwreck of 3 February 2026 exemplifies the acute tensions between Greece’s increasingly securitized border management and its obligations under EU law, the ECHR, and international maritime conventions such as the 1979 Search and Rescue (SAR) Convention. More specifically, off the coast of Chios island, a Hellenic Coast Guard patrol vessel collided with an inflatable boat carrying approximately 39 Afghan nationals, resulting in 15 deaths and 24 injuries, including 11 minors, and cases of miscarriage. All of the deaths were attributed to severe head trauma rather than drowning, per up-to-date autopsy reports, with survivors claim that the coast guard did not offer any prior warning or communication before ramming the migrant boat, contrary to official claims of the migrants’ speedboat initiating contact. Interestingly, the on boat cameras of the patrol vessel had been deactivated.   

This event parallels the 14 June 2023 Pylos shipwreck, Europe’s deadliest maritime migration tragedy, where over 500 lives (mainly Syrian, Pakistani, and Egyptian) were lost after the overcrowded trawler Adriana capsized, allegedly due to Coast Guard towing maneuvers following delayed rescue operation, despite prior distress alerts. The ongoing criminal proceedings in Greece have charged 17 Coast Guard personnel, including the rescue vessel captain, with felony offenses such as endangering lives and contributing to the shipwreck,. This development appears to be part of a systematic attempt to portray the eastern Mediterranean migration route as inherently life-threatening for asylum seekers, thus reflecting a pervasive securitization narrative guiding border policies of the Greek government that overshadows State accountability.

Legal Parallels and Accountability Gaps

From a legal aspect, both incidents implicate Greece’s positive obligations under Article 2 ECHR (right to life), requiring States to safeguard lives within their jurisdiction, including during maritime interceptions, and conduct effective, independent investigations into fatalities. The ECtHR has repeatedly held Greece accountable in analogous cases: in Safi and Others v. Greece, for inadequate protection and probing of a sunk migrant boat; Alkhatib and Others v. Greece, for excessive lethal force lacking “absolute necessity” and deficient regulatory frameworks for Coast Guard firearms use. Such repeated failures in border management operations seem to formulate a consistent pattern of action of the Greek authorities in handling migrant routes, in an attempt to not allow migrants to enter Greek territory/territorial waters. In this sense, although Article 3 ECHR (prohibition of inhuman/degrading treatment) further prohibits collective expulsions or pushbacks, this practice has been deemed systematic by Greek authorities in A.R.E. v. Greece (also see, here).

Under EU law, the Qualification and Asylum Procedures Directives respectively mandate the upholding of the principle of non-refoulement and individual assessments of asylum applications, while Article 4 of Protocol 4 of the ECHR and Article 19 CFR bar collective expulsions. Meanwhile, the SOLAS and SAR Conventions impose duties to render assistance “without delay” to persons in distress, disembarking them to a place of safety, irrespective of nationality or the existence of a right to enter the country. Hence, any form of interception framed as SAR mission cannot justify pushbacks or endangering the lives of migrants.

Greece’s Restrictive Policies and Criminalization of Solidarity

Domestically, this incident aligns with broader migration policy tendencies, as Greece has instrumentalized criminal law in an attempt to restrict migration, rendering irregular entry, stay and exit of the country a felony punishable with up to 5 years of imprisonment coupled with a minimum fine of €5,000 (Law 5226/2025 Government Gazette Α' 154/8.9.2025). Similarly, rejected asylum seekers face administrative fines up to €10,000, as well as up to five-year sentences or electronic ankle monitoring. Hence, the 2025 deportation law, hailed as Europe’s most stringent, essentially attempts to streamline expulsions of “economic migrants,” given that long-term regularization after 7 years of stay in the country is equally abolished.

This framework cannot be dissociated from Greece’s post-2019 migration hardening: escalated border fortifications (Evros 35 km wall), freezing of asylum applications, and systematic pushbacks exceeding 540 incidents between 2020-2022 (also see, here). Hence, high-seas shipwrecks, such as the Chios and Pylos lethal incidents, epitomize how this apparatus practically overrides positive obligations under the ECHR, as well as international humanitarian and maritime law, subordinating the protection of life at sea to national security imperatives.

The Greek Government defends its approach as prevention of illegal entry, invoking safe third country safeguards, especially for migrants arriving from Turkey, yet these yield no derogation from non-refoulement or collective expulsion bans. Hence, the Government’s approach in migration policies embodies a “fortress mentality,” which, coupled with its recent attempts to criminalize solidarity, further sets in danger the lives of migrants attempting to cross the Eastern Mediterranean route. A prominent example of this criminalization tendency include the recent Lesbos case against 24 rescuers, who were acquitted after years on charges, like espionage and smuggling, that carried up to 20-year sentences. Similarly, Norwegian activist Tommy Olsen faced an arrest warrant in February 2026 for documenting pushbacks via Aegean Boat Report, accused of criminal organization. Finally, a February 2026 migration law amendment (Law 5275/2026, Government Gazette Α’ 17/06-02-2026) makes NGO membership an aggravating factor, escalating misdemeanors (e.g., facilitation of stay) to felonies with fines exceeding €100,000, constituting the largest criminalization of solidarity in the EU.

Analysis

It is apparent that the Chios tragedy, when assessed alongside the Pylos shipwreck, does not constitute an isolated operational failure but rather indicative of a structural recalibration of border governance in the Eastern Mediterranean. The shift from the enforcement of search-and-rescue obligations to human rights violations at the EU’s external borders under SAR cover reveals a normative inversion: life-saving obligations are operationalized through a security prism that treats irregular entry primarily as a threat vector rather than a protection trigger.

At the doctrinal level, Article 2 ECHR imposes both substantive and procedural duties on States. More specifically, from a substantive aspect, States shall refrain from unlawful deprivation of life and adopt preventive operational measures where authorities knew or ought to have known of a real and immediate risk. Meanwhile, procedural obligations mandate to conduct prompt, effective, and independent investigations capable of leading to accountability. In both Chios and Pylos, the central legal question is whether Greek authorities fulfilled the due diligence threshold required during maritime interception. The reported deactivation of onboard cameras in Chios and the delayed rescue response in Pylos indicate the State’s unwillingness to comply with these operational standards.

In a similar vein, the 2026 legislative reform represents an internal consolidation of the securitization paradigm. By reclassifying irregular entry and facilitation-related conduct as felonies and elevating NGO affiliation to an aggravating factor, the Greek legal framework operationalizes criminal law as a migration-management instrument, fully adopting a “Crimmigration” approach, that is the convergence of criminal and immigration enforcement logics, as border management framework.

In this sense, the prosecution of humanitarian actors in the Lesbos case and proceedings against figures associated with monitoring networks reinforce a chilling effect on civil society oversight. When accountability mechanisms (NGO monitoring, documentation of pushbacks) are suppressed, the evidentiary architecture for fundamental rights protection is simultaneously weakened. In practical terms, criminalization of solidarity indirectly facilitates impunity.

Greece’s approach cannot be decoupled from the broader EU externalization strategy. Financial and operational support through Frontex, coupled with political endorsement of deterrence metrics (reduced arrivals as “success indicators”), generates structural incentives that privilege interdiction over protection. In this context, it appears that the Eastern Mediterranean has become a testing ground for this hybrid governance model of the EU’s external borders. This primarily includes operational opacity (restricted access, disabled recording systems), normative elasticity (expansive security justifications), and penal reinforcement (domestic felony frameworks). In other words, the legal tension at stake is not merely compliance with international human rights law but the hierarchy of values underpinning EU border management. If border integrity consistently supersedes the core values of life and human dignity, the doctrinal architecture of human rights law is functionally subordinated to security rationales.

Accordingly, the Chios incident should be analyzed not only as a maritime tragedy but as a constitutional stress test for the EU human rights regime. The decisive issue is whether accountability mechanisms, domestic courts, the ECtHR, EU oversight bodies, will be able to effectively recalibrate operational practice toward a life-preserving baseline or tacitly normalize deterrence-driven fundamental rights erosion.

Conclusion

Greece exemplifies a broader European paradigm: a so‑called “success story” for deterrence‑based migration control, yet in reality a humanitarian catastrophe for those seeking protection. The country’s migration policies mirror a wider EU strategy that prioritizes border fortification over human life. Hence, a rights‑first recalibration is urgently required. This entails independent and transparent investigations into all reported maritime incidents, such as the full public release of the Chios and Pylos footage, and unhindered support for NGOs engaged in SAR operations, paired with the domestic decriminalization of humanitarian assistance to migrants. In the absence of these measures, the prevailing doctrine of “prevention at all costs” will perpetuate watery graves, turning the Mediterranean into an open cemetery and rendering the protection of migrants’ fundamental rights mere eulogies in default.