Frontex’ Responsibility for Human Rights Violations: The CJEU and Certain Aspects of the International Responsibility of International Organisations

Setting the scene: Frontex, CJEU and the ARIO…and beyond

Although the Articles on the Responsibility of International Organizations (ARIO) were adopted 15 years ago by the International Law Commission (ILC), international judicial practice interpreting and unpacking the ARIO remains scarce. That the case law of regional courts can contribute to shedding light onto ARIO rules and concepts can be seen, inter alia, with regard to the East African Court of Justice (see the UN Secretary General’s compilations here and here). In a similar vein, it is worth turning to the Court of Justice of the EU (CJEU), notably its two recent Grand Chamber rulings delivered on International Migrants’ Day (18 December 2025). Both CJEU judgments concern compensation sought by individuals from an organ of a regional international organisation (IO): the European Border and Coast Guard Agency, Frontex. Frontex is the EU’s border management and return-enforcing ‘operational arm’. The individuals alleged human rights violations by Frontex in the course of its involvement in joint sea-border surveillance operations and a joint return flight, respectively. The two CJEU rulings serve to develop the law of responsibility of IOs, mostly beyond what is captured in ARIO.

WS and Others v Frontex concerns several Syrian nationals (including children) who, after having unsuccessfully sought asylum in the Greek islands in the Aegean in 2016, had been removed to Türkiye via a joint return flight co-organised and implemented by Frontex and Greece. They eventually moved on to Iraq, where they settled. In Hamoudi v Frontex, the applicant claims his fundamental rights were violated in 2020 when the Greek authorities pushed him and others back to sea on a life-raft in an area of the Aegean where Frontex aerial surveillance and information-sharing activities were ongoing. This prevented his right to seek asylum and threatened his life, amongst other alleged rights violations.

In both cases, the applicants posit: in the joint operations concerned, Frontex did not discharge its fundamental rights due diligence obligations under its Founding Regulation and engaged, directly or indirectly, in improper conduct with regard to them. They claim compensation for material and non-material damages suffered as a result of Frontex’ failure to comply with its fundamental rights obligations, activating the action for damages under Articles 268 and 340 of the Treaty on the Functioning of the EU – also enshrined in Article 41(3) of the Charter of Fundamental Rights of the EU (‘the Charter’). These fundamental rights obligations include the prohibitions of refoulement and collective expulsion, the right to asylum, prohibition of ill-treatment, and the right to an effective (judicial) remedy. The EU General Court, acting at first instance, had dismissed both claims. These orders were subsequently appealed before the Grand Chamber. After the Grand Chamber quashed the initial decisions, both cases are now back before the General Court for a repeat procedure.

These landmark CJEU Grand Chamber rulings form part of the long saga of attempts to hold Frontex responsible for direct or indirect violations of fundamental rights of people on the move. They have already been subject to EU-law centred commentaries in the legal blogosphere (see Tas, Callewaert, Pirrello and Nicolosi, Ziebritzki, Kunst, Schubert, and Pirrello and Aviat). More (European law) analysis will surely blossom. This post takes a different angle by adopting an international law lens. It aims to distil those elements of the two CJEU judgments which, although not expressly mentioning the ARIO, can contribute to illuminating certain aspects of the shared or parallel responsibility of IOs for human rights violations in general. These primarily concern making reparation for injury caused by the internationally wrongful conduct of IOs, such as issues of causation, burden of proof and standard of proof.

IOs’ human rights obligations and ARIO’s relevance when making good the injury caused by human rights violations

It is well known that the responsibility of IOs extend to human rights violations – Articles 3-4 ARIO imply it, too. IOs may, and do, have human rights obligations. They are principally based on customary international law or even general principles of law (Alston and Simma, Faix, Johansen) or as some argue, derived from their Member States’ relevant treaty obligations (Ahmed and Butler). Such obligations can be also incorporated into their constituent instruments/secondary law, coupled with the very rare case of an IO becoming party to human rights treaties (a case in point is the EU). An implicit recognition of IOs as duty-bearers under international human rights law can also be found in Article 53 ARIO (countermeasures must not affect IOs’ human rights obligations). Parts One and Two of ARIO can thus kick in whenever an IO violates its human rights obligations.

Part Three of ARIO is ‘without prejudice to any right’ directly resulting from the breach of obligations owed to non-State entities (e.g. individuals) under international law (Article 33(2) ARIO). Nevertheless, pure logic, the unity of law and other compelling reasons, such as the general lack of human rights adjudication mechanisms to which IOs are party, strongly suggest applying ARIO rules governing reparation in cases concerning violations of individuals’ human rights, too. By the same token, the International Court of Justice long ago indicated, in a 1973 Advisory Opinion, that the basic principle of reparation also applies to reparation for injury to individuals. There is thus nothing against conceiving reparation-related basic concepts as common across the entire body of international law.

Some underdeveloped secondary norms of primary importance and the CJEU’s arrival to the party

Despite the above conclusion about the applicability of ARIO concepts to human rights breaches, several aspects of making good human rights infringements by IOs (e.g. compensation) still remain underdeveloped in the ARIO and legal scholarship, not to mention the inconsistencies and contestations in international jurisprudence. Hence the added value of the two CJEU Grand Chamber pronouncements in relation to some of these underexplored aspects requiring further illumination and judicial unpacking.

Concurrent responsibility

The repeated examination of the cases by the General Court – following the Grand Chamber’s guidance – can also usefully contribute to better grasping how shared or concurrent responsibility of IOs and States actually work in practice. It is clear from the CJEU’s assessment in WS and Others that Frontex breached its own fundamental rights obligations under EU law. The Agency did so by not verifying the existence of an enforceable return decision and may have breached more by contributing to sub-standard treatment during the joint removal flight and not performing its monitoring duties (paras. 102, 132). As a result, the scenario of Frontex as a ‘bystander’, merely aiding or assisting in the commission of an internationally wrongful act by a State (Greece) within the meaning of Article 14 ARIO does not come into play here.

Given its direct breach of law – and other possible infringements – independently from and next to the Greek authorities, Frontex may have thus incurred concurrent (parallel) responsibility with the Member State concerned. This enters the scenario described in Article 48 ARIO: where one IO and one or more States are responsible for the same international wrong (a situation that I have already contemplated in a previous post on this blog). Irrespective of whether the claimed joint and several responsibility of Frontex will be eventually established in casu (something that the Grand Chamber has not ruled out in principle – see Kočner v Europol and more broadly the Guiding Principles on Shared Responsibility in International Law), we are in a situation of a combination of wrongdoers, yielding a sort of shared or concurrent (parallel) responsibility. Its concrete form, including whether such wrongful conduct entails Frontex’ joint and several responsibility for the damage caused, will have bearings on apportioning the exact amounts of monetary compensation. This exercise will also need to take into account the sums the ECtHR already awarded in an amicable settlement with Greece for the applicants’ damage (WS and Others, para. 35).

Causation

This brings us to the next question in the logical chain of legal issues under scrutiny. It is the existence of a causal link – and its directness – between Frontex’ wrongful conduct and the damage suffered; as well as the related question of how the causal link can be broken. Regarding the applicants’ decision in WS and Others to move on to Iraq from Türkiye, the CJEU held: while their own conduct can, in theory, break the casual link, its examination must be carried out on a case-by-case basis, taking into account individual circumstances ‘characterising the situation of the adversely affected person’ (para. 154). The Grand Chamber also emphasised the vulnerability of asylum applicants. They, as human beings facing ‘unforeseeable risks’ and possibly exposed to ‘traumatic experiences’, cannot be required to make purely rational decisions in the same way as economic operators do (paras. 156-157) – or other IOs with their complex structures and professional decision-making processes. This is an important contribution to the embryonic causation concept in Articles 31, 34 and 36 ARIO. It equally helps context-specifically lower the otherwise controversial causation tests in international courtrooms, to be driven by such ‘reasonable response’ criterion when individuals seek compensation for IOs’ human rights violations. 

Burden and standard of proof

Another relevant aspect pertains to the burden of proof where the Grand Chamber implicitly recognized the possibility of reversing it. In the words of the EU Court in Hamoudi, ‘failure to adapt the burden of proof … might hinder all legal action by victims of a pushback operation against Frontex on the basis of [its] alleged unlawful conduct …, granting [it] de facto immunity and thus jeopardising the effective protection of the [victims’] fundamental rights’ (para. 105). What is clear is that the burden of proof is at least shared in such situations. Frontex is bound to provide all relevant information in its possession (e.g. from the footage of the aerial surveillance aircraft) to ascertain whether the pushback at sea has actually taken place and if the applicant was present.  

A connected aspect that the CJEU dwelt upon is the required standard of proof. Drawing on ECtHR case law, the Grand Chamber held in Hamoudi that in view of giving full effect to the right to effective judicial protection (Article 47 of the Charter) and not to make it illusory, pushback cases warrant an adaptation of the standard of proof (paras. 104-112). This means that alleged pushback victims cannot be requested to demonstrate conclusive proof, and the pieces of evidence adduced by the applicant in the case at hand were sufficiently detailed, specific and consistent to constitute prima facie evidence. When opting for this reasonably lowered standard of proof, the Grand Chamber recalled the principle of the unfettered assessment of evidence. Under this, the only relevant criterion for the purpose of assessing the probative value of evidence is its credibility (para. 114). The right to an effective judicial remedy, recognised as part of customary international human rights law and enshrined in numerous universal and regional human rights treaties, has been used as a legal hook to argue for a lower standard of proof and the possibility of reversing the burden of proof. Strategically mobilising the human right of access to justice presents yet another promising opening to more effectively hold IOs accountable for human rights violations.

As a common thread in both cases, acknowledging the applicants’ vulnerability as asylum seekers can be a tool for other adjudicators and similar bodies to apply the proper legal tests concerning the causal link as well as the burden and standard of proof. This with lower thresholds and sufficiently factoring in the human element. Such an approach would be a desired step forward to efficiently address the factual power imbalances tilting towards IOs and the weak position of the individual towards them.

Outro – with an eye on the future

Ultimately, this brief analysis aimed to offer some insights into how the two recent CJEU Grand Chamber cases on holding Frontex responsible for its human rights violations in multi-actor cooperative migration management scenarios can contribute to elucidating the responsibility of IOs for human rights violations more generally — in particular the issues of shared responsibility and compensation-related procedural rules. This is not purely theoretical. IOs are remarkably active, including in the field of migration – consider, for example, the UN Refugee Agency managing refugee camps, and the International Organization for Migration co-organising, as a global service-provider, irregular migrants’ ‘assisted voluntary returns’. Also, the CJEU findings can meaningfully inform two strands of the International Law Commission’s ongoing work. One is the settlement of disputes to which international organisations are parties (human-rights related disputes between IOs and individuals represent a prominent part of IOs’ dispute-settlement practice). The other is compensation for the damage caused by internationally wrongful acts, including those committed by IOs. Besides various judicial and non-judicial international dispute settlement bodies, eyes are now on the EU General Court’s second assessment. The EU judges acting again at first instance may likewise engage with certain concepts of international responsibility of IOs – to showcase yet another area where the ‘legal convictions’ of both the EU and the international legal orders converge.