In the evolving landscape of EU border accountability, the Court of Justice of the European Union (CJEU) delivered its ruling in Hamoudi v Frontex (C-136/24 P) on 18 December 2025. This Grand Chamber decision not only reversed the General Court’s (GC) dismissal but also fundamentally recalibrated the evidentiary standards for establishing Frontex’s liability for fundamental rights violations, as discussed here. As FM v Frontex (T-511/24), a closely related case, awaits its judgment, Hamoudi’s legacy in establishing Frontex’s accountability and lowering the burden of proof for vulnerable migrants, promises to reshape its outcome. This post initially analyzes Hamoudi’s doctrinal mechanics, draws parallels to FM, and sketches plausible trajectories for the latter, thereby illuminating Frontex’s emerging liability regime. The central claim advanced here is that the Court in Hamoudi revolutionized EU non-contractual liability under Article 340(2) TFEU by easing evidentiary burdens for migrants, shifting proof obligations to Frontex, and affirming the agency’s independent fundamental rights duties. In doing so, it sets the stage for the pending case FM v Frontex on aerial surveillance and data-sharing complicity.

The Doctrinal Foundation: Union Non-Contractual Liability Under Article 340(2) TFEU

Before examining Hamoudi, it is useful to recall the framework governing EU Agency liability. Article 340(2) TFEU establishes that the Union shall make good any damage caused by its institutions or agents in the performance of their duties. CJEU case law has consistently held that three cumulative conditions must be satisfied: (1) an unlawful act, (2) a sufficiently serious breach of EU law, and (3) a direct causal link between the breach and the harm suffered. The substantive rules conferring rights. In this context, Hamoudi’s significance lies partly in how it recalibrates the evidentiary threshold for proving these conditions, particularly in factual circumstances where applicants lack privileged access to state or Agency documentation.

Hamoudi v Frontex: A Watershed on Evidentiary Burdens

Mr Hamoudi, a Syrian national, alleged Frontex complicity in his 2020 illegal pushback from Greek territorial waters to Turkey. Despite video evidence and NGO reports implicating Frontex air patrols, the GC (T‑136/22) dismissed the claim for lack of sufficient causal proof establishing the Agency’s accountability. The Court annulled this judgment, articulating three substantive and interrelated shifts:

Prima Facie Evidence Suffices

Migrants need only present “serious indications,” such as videos, Frontex operational logs, witness testimony, and corroborating NGO documentation, to establish prima facie Frontex involvement and causation of harm. The Court grounded this reduced threshold in the principle of effective access to justice under Article 47 CFR (right to an effective remedy). Critically, it would be disproportionate to demand full documentary proof from asylum seekers who, by definition, lack access to classified Frontex operational data, surveillance recordings, and internal communications. The burden shifting thus rests on an asymmetry of information: only state authorities and Frontex possess the documentary evidence necessary to construct a complete causal narrative.

Burden Shift to Frontex: Non-Disclosure as Adverse Inference

Once a migrant establishes prima facie evidence meeting the threshold of “serious indications,” the evidentiary burden shifts. Frontex is then obliged to come forward with full disclosure of internal reports, mission logs, operational data, and communications that either rebut or clarify the alleged involvement. Significantly, the Court held that non-disclosure or silence, without compelling operational security justifications, hampers the Court’s liability assessment. This approach is grounded in Article 47 CFR and the principle of access to effective remedy: a party cannot simultaneously benefit from institutional opacity and deny liability arising from that opacity. In this context, non-responsive silence substantially weakens Frontex’s defensive position and may result in tipping the evidentiary balance in applicants’ favor.

Independent Obligations: Frontex as More Than a “Mere Executor”

The Court further rejected the characterization of Frontex as a mere executor of Member State decisions. Instead, the Court held that Frontex bears independent, non-delegable obligations to assess the lawfulness of operations before, during, and after execution. Specifically, Frontex must independently verify compliance with the principle of non-refoulement (Article 19(2) CFR) and the right to asylum (Article 18 CFR) before it engages in operational support. This duty is not discharged by virtue of Member State authorization; Frontex retains an autonomous responsibility under EU primary law to refuse participation in manifestly unlawful operations. A breach of this duty (for instance, using technical or operational support to known pushback or pullback schemes) constitutes an unlawful act under Article 340(2) TFEU.

FM v Frontex: Aerial Surveillance and Data Sharing as Alleged Complicity

In October 2024, front-LEX and Refugees in Libya brought an action for failure to act against Frontex before the GC on behalf of FM, a Sudanese asylum seeker trapped in Libya without access to international protection or medical care. The applicant contends that he faces heightened risk of detection if he attempts to cross the Mediterranean, consequent handover to Libyan entities by intercepted vessels, and subsequent serious abuse including detention and torture. He seeks a declaration that Frontex’s unlawful information sharing violates Article 46(4) of the Frontex Regulation (Regulation 2019/1896) and demand suspension or termination of the Agency’s geolocation data sharing with Libyan authorities.

The applicant alleges that Frontex disseminates real time geolocation data regarding refugee boats traversing the Central Mediterranean to Libyan “rescue” coordination centers and other Libyan entities. This data sharing systematically facilitates the interception and pullback of migrant vessels to Libya, where FM and others face serious risks of persecution, torture, arbitrary detention, and inhuman treatment. Although one might argue that Frontex’s sole obligation is to inform the internationally recognized responsible rescue coordination centre, the principle of non-refoulement bears broader duties for Frontex not to facilitate returns to places of persecution.

Furthermore, the applicant had previously submitted a legal notice under Article 265 TFEU (action for failure to act), demanding that Frontex’s Executive Director suspend or terminate the Agency’s aerial surveillance in the Central Mediterranean “pre-frontier area” pending a human rights audit. He argued that the Executive Director failed to provide adequately reasoned justification for rejecting their request to act (dated 29 May 2024) and did not clarify the Agency’s position in response. It has to be noted that, in April 2025, the GC declined to dismiss FM’s complaint as inadmissible and instead postponed the inadmissibility ruling pending examination of the case on the merits. This procedural step signals that the Court will engage substantively with FM’s factual allegations and not shield the Agency behind standing or justiciability barriers.

Drawing Parallels: Hamoudi’s Logic Applied to FM

The three main strands of Hamoudi’s reasoning (reduced evidentiary thresholds, burden shifting and independent Agency obligations) are directly transferable to FM’s legal issues. Hence, the GC in FM will be bound by Hamoudi’s Grand Chamber pronouncements and should follow the same approach. This means: (i) FM’s corroborated allegations of Frontex surveillance, geolocation data sharing, and complicity in Libyan interceptions will satisfy the prima facie threshold more readily than what the GC was willing to accept in Hamoudi; (ii) Frontex will be obliged to disclose classified surveillance feeds, data-handling protocols, and coordination records with Libyan authorities; and (iii) Frontex cannot invoke Member State authority or third-country liability as a shield against its own independent duty to assess whether geolocation data sharing violates the principle of non-refoulement.

Hence, FM’s claim rests on establishing a sufficiently serious breach of Frontex’s independent obligations under the principle of non-refoulement and the Frontex Regulation. Here, the causal chain is plausible: Frontex surveillance detects migrant boats; Frontex transmits geolocation data to Libyan authorities; Libyan entities intercept the vessels; migrants are returned to Libya; FM faces persecution and torture. The critical question is whether Frontex’s participation in the first two steps (surveillance and data sharing) bears sufficient causal responsibility for the harm suffered in steps three and four, given that Libyan authorities and intercepting vessels are intervening actors.

Why Causation Is Plausibly “Direct Enough”

Under Article 340(2) TFEU, causation does not need to be sole causation. Hence, contributory causation suffices. Where an Agency’s conduct (data sharing) substantially facilitates a foreseeable harmful outcome (unlawful detention and ill-treatment in Libya) despite the presence of intervening third country actors, Agency liability should be recognized. The Hamoudi judgment emphasized the Agency’s obligation not to participate in manifestly unlawful operations, thus strengthening this causal connection: if Frontex acts knowing, or reasonably should know, that geolocation data will enable Libyan interceptions and returns to persecution, then Frontex’s conduct is not merely correlative but actively facilitative of the harm. Consequently, the causal link is “direct enough” to satisfy Article 340(2) TFEU requirements, notwithstanding Libyan authorities’ intervening role.

Application of Hamoudi’s Evidentiary Standards

Post-Hamoudi, FM should be able to establish prima facie evidence through, for example, NGO documentation of boat interceptions in the Central Mediterranean, corroborated testimony from migrants describing Libyan interception and abuse, or reports from human rights organizations linking interceptions to Frontex surveillance patterns and timings. These elements, individually, may not constitute ironclad proof, but they may meet the threshold of “serious indications” sufficient under Hamoudi.

Once FM meets this prima facie threshold, the evidentiary burden shifts. Frontex will have to produce operational logs, surveillance flight records, data sharing protocols, and communications with Libyan entities demonstrating either that no geolocation data was shared, or alternatively, that such sharing was compliant with non-refoulement obligations or sufficiently removed from the ensuing harm.

Two Plausible Judicial Pathways

The GC in FM faces two broadly distinct doctrinal choices, each internally coherent:

Pathway 1: Robust application of Hamoudi

The Court accepts FM’s prima facie evidence as sufficient to trigger Hamoudi’s burden shifting mechanism. Frontex is ordered to disclose surveillance records and data sharing protocols. The Court finds that Frontex knew or should have known that geolocation data would facilitate Libyan interceptions. The Agency’s independent obligation under non-refoulement was breached because it participated in an operation with a foreseeable and serious risk of facilitating returns with persecution risks. The Court awards a declaration of unlawful action (or failure to act) and possibly damages.

Pathway 2: Narrower application

The Court accepts Hamoudi’s framework in principle but applies it more restrictively. It requires FM to present more granular evidence directly linking specific geolocation transmissions to FM’s own detention and harm (as opposed to general patterns of interception). It permits Frontex greater deference on operational security grounds and accepts the Agency’s argument that responsibility lies primarily with Libyan entities and Member States. It may find narrow liability for specific, demonstrably unlawful data sharing instances but decline broader findings of systematic breach.

Thus, the outcome hinges on how the GC will apply the Hamoudi judgment. Should the GC opt for the first pathway, this will assist in solidifying protection of migrants fundamental rights at the EU’s external borders and beyond. On the contrary, if the GC follows the second pathway, the Agency’s institutional flexibility and operational scope would remain largely intact, on the cost of robust fundamental rights protection.

Conclusion

The Grand Chamber’s ruling in Hamoudi v Frontex marks a significant shift toward holding EU Agencies accountable for fundamental rights breaches by dismantling evidentiary protections and reinforcing transparency obligations. By lowering the initial burden of proof and shifting the evidentiary load to Frontex once prima facie evidence is presented, Hamoudi substantially facilitates liability findings in opaque, asymmetric factual settings. FM v Frontex constitutes the first test on whether Hamoudi’s logic extends to extraterritorial surveillance and data sharing complicity in third country abuses. The GC’s decision to proceed to the merits in FM, rather than dismissing on procedural grounds, suggests institutional willingness to engage substantively with Agency accountability questions. Post-Hamoudi, the legal foundation for holding Frontex liable exists. Now it is up to the GC to deploy its potential. This will be dependent on how it will calibrate the competing interests of border security, institutional autonomy, and protection of migrants’ fundamental rights. Hence, the stakes are high: FM’s outcome will shape the external boundaries of Frontex’s liability regime and signal whether EU Agencies can ultimately evade responsibility for systematic participation in practices in extra-EU territory that risk violating the principle of non-refoulement, which constitutes one of the cornerstones of the Union’s constitutional commitment to fundamental rights.