The Frontex Doctrine: Concurrent Responsibility Beyond Articles on the Responsibility of International Organisations

[Arko Sankar Karmakar is a third year law student from the West Bengal National University of Juridical Sciences, Kolkat]

Introduction

Debates on the responsibility of international organisations have been mainly influenced by the ILC’s Articles on the Responsibility of International Organisations. ARIO has never been in an easy spot. The text is detailed and it is normatively weak and institutionally lacking. It  negative criticism, because of the abstract manner in which it addresses issues such as concurrent responsibility, causation, and evidentiary burdens in scenarios where states and international organisations act jointly. It has hardly generated any solid judicial practice. By contrast, the recent judgments of the Court of Justice of the European Union in WS & others and Hamoudi v Frontex mark a turning point. The Court, for the first time, identified an IO as responsible for human rights violations that occurred during joint operations at the EU’s external borders.

This post argues that the “Frontex doctrine” does more than develop EU liability law. It constitutes a judicially-crafted regime of concurrent responsibility for international organisations that is functionally equivalent to general international law rules, but is sourced in EU law and backed by a dense enforcement apparatus. In doing so, it exposes the limitations of the ARIO framework. It illustrates a practical form of “regime displacement” not in the sense of formal replacement, but as a shift in practical authority. Regional judiciary has overtaken and marginalised the ILC project as the operative reference point for IO responsibility in joint operations.

ARIO and the Problem of Joint Operations

The ARIO were designed to transpose the logic of state responsibility to international organisations. However, when it comes to joint operations, now the default mode in fields such as border control, peacekeeping, and migration management, their guidance is strikingly sparse. The provisions that “aid or assist” a state, or that circumvent its obligations by using states as instruments, are heavily modelled on the ARIO.

The Articles codify existing customary international law and create new rules. The ILC itself recognised that it is not always easy to identify the two elements separately. In contrast to the earlier Articles on State Responsibility, which international tribunals have extensively cited as authoritative restatements of custom, ARIO have generated very little judicial practice, and their claim to represent lex lata remains a matter of debate. Michael Wood has identified a few criteria to determine the degree to which ARIO correspond to customary international law. In contrast, scholars have discussed whether certain articles of the text, especially those related to joint operations and concurrent responsibility, reflect changing practice or merely express desired norms. This lack of clarity regarding ARIO’s status is key to understanding its practical limitations and the importance of parallel developments in regional orders.

Three weaknesses are particularly acute:

First, ARIO did not develop concrete criteria for when concurrent responsibility arises in complex, multi-actor operations. It recognises the possibility that both an international organisation and one or more states may be responsible in parallel. Second, ARIO assumes a relatively linear relationship between conduct and injury, which is ill-suited to situations in which different actors plan and execute a joint operation over time. Third, the ARIO assumes victims need to prove wrongdoing, even though they often don’t have all the information that big companies do.

WS and Others v Frontex: From Support Actor to Responsible Organ

In WS and Others, a family had sought asylum in Greece and were subsequently removed on a Frontex-coordinated return flight. They alleged that Frontex had failed to ensure that valid return decisions existed, to safeguard access to the asylum procedure, and to prevent inhuman treatment during the operation, in breach of EU law Article 340(2) TFEU sets out the rules governing the EU’s non-contractual liability. This provides that the Union’s staff who are carrying out their duties are responsible for making good any harm caused by them, in accordance with the general principles common to the laws of the Member States. To be able to prove the fault of the Member, it is necessary that three conditions have been met: The EU institution must have committed an unlawful act; the applicant must have suffered real damage; and there must be a link of causality between the unlawful act and the damage. This set of rules also applies to EU agencies such as Frontex. The Court’s power to hear cases for damages against EU bodies is derived from Article 268 TFEU, with the General Court being the court of first instance for claims brought by individuals.

The General Court dismissed their action on the basis that Frontex was only present to support through technology and handle day-to-day operations, that it is the country’s responsibility to ensure that rights are respected, and that they could not prove that Frontex caused the damage.

However, on appeal, the CJEU annulled this limited interpretation. It basically made three essential steps towards the Frontex doctrine.

• Re-characterising Frontex’s role: The Court found that portraying Frontex as a purely auxiliary actor was unacceptable and, thus, the Agency’s responsibility to ensure that only persons subject to enforceable return decisions are placed on such flights.
• Opening the door to concurrent responsibility: The Court considered that the General Court made a legal error by deciding that violations of fundamental rights during a joint operation are to be regarded as the sole responsibility of the Member State. On the contrary, the Court allowed that both Frontex and the state may be held liable if their respective roles significantly contributed to the harmful result.
• Reconsidering causation in joint operations: The Court condemned the General Court’s inflexible approach to the concept of causation and stressed that the analysis must be in concerto and thus should not be simplified into a formalistic division between the plan and the execution.

Put simply, these measures together convey a duty to check legality and give a fuller picture of the contributory role in joint operations. In a way, they are, in fact, providing legal backing for the idea that a common area of responsibility.

Hamoudi v Frontex: Adapting Burdens of Proof

WS and Others focuses on the substantive side of concurrent responsibility, Hamoudi addresses the evidentiary dimension. The case concerned an alleged pushback involving Frontex in the Aegean. Mr. Hamoudi sought damages, but the Court dismissed his action, finding that he had not conclusively proved his presence at the pushback and refusing his request that Frontex be ordered to produce operational documents in its possession. Since Frontex itself provided vital data, the Court of Justice held a conflict between the right to a fair trial, the burden of evidence and a situation where Frontex itself provided vital information. The Forum stresses that Frontex, as part of the European Border, is entrusted with collecting operational statistics and safeguarding essential immunities during its operations. Secondly, in the event of a perceived rejection, the requirement that candidates provide sufficient evidence of their attendance would, in practice, grant Frontex de facto immunity, given the real difficulties of obtaining testimony close to the sea and in border districts.

The CJEU therefore required an adaptation of the burden of proof. Claimants need only present prima facie evidence that a pushback occurred and that they were present; courts must then order Frontex to produce relevant information in its possession. The failure to explore such evidentiary avenues renders judicial protection illusory.

The current move is a departure from the general position. While formally couched in EU law, it is difficult to ignore its resonance with ongoing debates about adapting evidentiary standards in international and regional human rights fora to guarantee access to justice for migrants and refugees.

Functional Equivalence and the Question of Custom

The EU is not a party to any treaty that codifies ARIO, and the CJEU clearly bases its reasoning on the Frontex Regulation, and the general principles governing the Union’s non-contractual liability. ARIO, by contrast, concerns inter se responsibility between subjects of international law, yet the two regimes are functionally similar in allocating responsibility in joint operations.

These regimes take effect at different normative levels and pursue different functions. ARIO addresses international responsibility and establishes intersubject accountability. However, EU non-contractual liability is a tool for individual victims to get a remedy within the EU. It is a form of administrative accountability stemming from the EU’s constitutional nature. Notwithstanding these differences, the two systems are, in one specific respect, functionally equivalent; both determine responsibility when states and international organisations are involved in joint operations.

The Frontex ruling does three things that ARIO aims at but does not realise:

• It signifies that an international organisation could be concurrently responsible with states.
• It interprets the issue causally, meaning that the organisation’s failure to use those powers can be a legally relevant factor in the injury.
• It modifies the rules of evidence in accordance with the constitutional realities of international organisations by shifting the burdens and revising the standards of proof in favour of the claimants.

This functional equivalence, however, must not be taken as the same action. Since the ARIO and EU rules on non-contractual liability belong to different levels, the Frontex legal principle does not override the ILC concept. The relationship between the two is best understood not as substitution but as parallel development, in which EU liability law borrows and operationalises certain ideas associated with ARIO without purporting to replace general international law. The two systems coexist, but in the context of joint border operations within the EU, where individuals seek justice, the EU liability system has been the primary framework, leaving ARIO as a secondary source.

This leads us to a more acute question: if regional courts set such standards, are they effectively contributing to customary international law, or are they forming a separate regime that stands in opposition to it?

According to one interpretation, the Frontex doctrine is a kind of state practice accompanied by opinio juris: Member States and the EU, in their positions in litigation and regulatory reforms, acknowledge that an IO which is practically empowered to carry out the joint operations must ascertain the legality, may be held responsible in case of violations, and must cooperate in evidentiary matters. If such behaviour stabilises and later spreads through citations in other regional courts and doctrinal uptake, it might become customary rules on concurrent responsibility and burden shifting in IO contexts.

Another view suggests that CJEU criteria constitute a parallel regime that is self-contained and robust. EU law derives its legitimacy from the Union itself, and the binding character of these judgments depends on their internal judicial architecture rather than on international dispute settlement. Yet, their effective, victim-oriented enforcement, confronts the under-enforced and non-binding ARIO framework without purporting to replace it as general international law. Therefore, the Frontex doctrine does not displace ARIO as a matter of formal doctrine, but it does reduce ARIO’s practical centrality for questions of IO responsibility in EU joint operations.

Regime Displacement and the Future of IO Responsibility

The relationship between the Frontex doctrine and the ILC’s ARIO is therefore not one of straightforward replacement. According to one interpretation, the ILC’s ARIO project, which was initially aimed at reflecting and developing general international law, is now being supplemented and, in some respects, overshadowed by court-driven interpretations of IO responsibility grounded in regional orders. Another interpretation holds that these regional developments do not merely displace ARIO but also engage with and operationalise certain principles relating to the international responsibility of international organisations within a much stronger institutional framework. The EU litigation can be seen as indirectly reinforcing core concepts from ARIO, such as concurrent responsibility, the relevance of organisational powers in determining liability, and the attribution of conduct, while developing them within the distinct architecture of EU law. The EU litigation can be seen as indirectly reinforcing core concepts from ARIO, while developing them within the distinct architecture of EU law.  One may argue that this dynamic is best described as parallel development with points of interaction: CJEU case law draws on concepts familiar from ARIO, but deploys them within the specific institutional and remedial context of EU law rather than as a restatement of general international law. These interpretations borrow the conceptual vocabulary of general international law but reframe it in the context of practical issues such as pushbacks and joint return operations and within institutional settings that can credibly provide remedies.

To begin with, it undermines the straightforward narrative that general international law on IO responsibility will be primarily developed through ILC drafting and traditional interstate practice. In fact, regional judicialized regimes may set de facto standards. Secondly, it highlights the significance of accountability and sensitive design. One may interpret ARIO’s reluctance to engage in causation and evidence as a political choice that sacrificed victims’ access for conceptual neatness. The Frontex doctrine is shifting the focus back to victim access.

Thirdly, it makes us think again about the relationship between EU law and general international law, rather than viewing EU law as an isolated system.

Conclusion

The CJEUs Frontex case law represents more than just a narrow episode of EU administrative liability. It creates a highly sophisticated and legally binding theory of concurrent liability for an IO engaged in joint operations, thus addressing shared responsibility, causation, and evidentiary burdens that ARIO left most under-theorised.WS places Frontex as a duty-bearing entity with its own verification obligations and the possibility of parallel liability; Hamoudi reallocates the burden of proof to prevent de facto immunity in an asymmetric evidentiary environment.

This “Frontex doctrine” should be seen as a system of IO liability essentially on the same level and capable of implementation within the EU legal order. Whether it becomes part of customary international law or remains a strong parallel regime, it effectively supplants ARIO as the primary point of reference for IO liability in joint operations, marginalising the ARIO framework in practice without formally replacing it. Lawyers cannot simply treat the IO’s responsibility as a dead ILC relic; rather, they must deal with the evolving, court-based norm-making at the EU’s external borders.