The creation of the European Anti-Money Laundering Authority (AMLA) in 2024 marks not just the latest chapter in EU agency proliferation, but also a moment to reflect again on the framework that was meant to guide the agencification process: the 2012 Common Approach on EU Decentralised Agencies. Intended to bring rationality to the sprawling landscape of EU agencies, the Common Approach already was an outdated “lowest common denominator” when adopted in 2012. With AMLA, the EU institutions have now moved well beyond it — sometimes for the better, but not always coherently.

This blogpost summarizes the main findings of a Chapter that is part of a forthcoming book edited by Jonathan Bauerschmidt, Raffaele D’Ambrosio, Diane Fromage and Andreas Schirk entitled AMLA and EU Financial Agencies to be published by Oxford University Press. It highlights three key issues presented by the newly created AMLA: the agency’s seat, its political accountability, and the legal protection mechanisms tied to its decision-making powers.

The AMLA Seat Debacle: A Governance Blind Spot

The selection of an EU agency’s seat has long been a politically sensitive issue, and AMLA proved no exception. The Common Approach had attempted, with some ambition, to bring structure to this process. It proposed several criteria and stressed the desirability of a geographical spread in EU agencies, while formally confirming that it was a prerogative of the Member States to select the seats of EU agencies “by common agreement.”

In the cases (1, 2, 3) on the seats of the European Medicines Agency and the European Labour Authority, however, the Court of Justice corrected this assumption and confirmed that as a matter of EU primary law, it is up to the EU legislator to select the seat of an EU agency, since it is also the legislator that decides to establish the agency. The AMLA subsequently was the very first agency the seat of which had to be selected in light of this judicial clarification. The process of doing so unfolded in an improvised and institutionally awkward manner. Parliament and Council each received 27 votes, but this attempt at institutional symmetry, at the insistence of Parliament itself, paradoxically left it worse off than under a standard consent procedure. Unable to veto a decision and faced with a Council which is typically able to maintain greater unity, Parliament was left politically sidelined, and the AMLA seat ultimately went to the Member States’ preferred pick.

Future reform could take inspiration from this misstep. One practical solution would be for the Commission to propose the seat location in its legislative drafts, a question which it now typically leaves blank in its proposals, following a transparent selection process. This would anchor the choice politically, allow for proper debate, and better reflect the role of the institutions in shaping agency governance. In any event, the institutional practice post-Common Approach, in which greater attention is being afforded to a seat selection based on objective criteria and following an assessment by the Commission should ideally be included in an updated Common Approach, even though the institutions do not seem inclined to revisit their inter-institutional agreement.

AMLA’s Political Accountability: A Structural Innovation

One of the most significant novelties in the AMLA Regulation lies in how it reconfigures the agency’s political accountability. Traditionally, the Common Approach framed agency directors as accountable to their Management Boards – that are mainly composed of national representatives  and, indirectly, to Parliament and Council via the annual discharge procedure. But this model was always incomplete. It left unresolved the question of who is accountable for strategic decisions taken by the Management Board — especially when those decisions have no budgetary implications.

AMLA marks a departure from this model. Its Management Board (called General Board in the AMLA Regulation) is (only) responsible for adopting the relevant supervisory measures and the draft delegated and implementing acts for the Commission. Like other agencies, the AMLA also has an Executive Board but its function is different from that of the other agencies’ executive boards. Where the typical executive boards supports the management board and prepares its work, the AMLA Regulation, indicates in its recital 61 that the Executive Board, rather than the Management Board is to be the AMLA’s governing body: “the Executive Board should be in charge of the planning and execution of all tasks of the Authority except where specific decisions are expressly allocated to the General Board”. The Executive Board thereby has a direct line of accountability to both the Council and Parliament. The AMLA Executive Board is composed of independent experts, selected through a multi-step appointment process set out in Article 63(4). Thus the Commission shortlists candidates, the General Board proposes them and the Council appoints them, subject to the approval of the Parliament. In case a member of the Executive Board no longer fulfils the conditions required for the performance of that member’s duties or has been guilty of serious misconduct, they can also be removed from office by the Council, acting on its own initiative or following a proposal from Parliament or the General Board.

This design whereby an agency is headed by a smaller committee of independent non-partisan experts brings AMLA closer to the model of independent regulatory agencies familiar in the United States. However, since the Executive Board members are not merely removable ‘for cause’, the AMLA is still not to be entirely equated with such US independent agencies. More importantly however, this new design enables—for the first time—an almost complete  accountability chain from the actors that ‘call the shots’ in the EU agency to the politically legitimate institutions. If Parliament or the Council is dissatisfied with AMLA decisions that are adopted by the Executive Board and not by the General Board, they can not only demand explanation but also initiate dismissal procedures in accordance with Article 63(7). In this sense, AMLA sets a potential blueprint for a new generation of EU agencies. Also this post-Common Approach innovation merits to be included in an updated Common Approach.

Administrative Review and Boards of Appeal: Continued Experimentation

The third domain where AMLA reveals institutional evolution — but also confusion — is in its legal protection framework. AMLA belongs to the subset of “decision-making agencies,” meaning it can impose binding individual or general decisions on third parties. Like other such agencies, it features an internal review mechanism: the Administrative Board of Review (ABoR) set up under Article 72.

The AMLA ABoR points to a deeper and still unresolved issue: the legal identity and role of these administrative review bodies, commonly called Boards of Appeal. The Common Approach mentions them only in passing, treating them on par with internal advisory committees rather than quasi-judicial actors. This ignores both their functional importance and the institutional consequences of recent legal developments.

Notably, the CJEU has clarified that Boards of Appeal are not ordinary internal units. In Aquind v. ACER, the Court described them as independent, quasi-judicial bodies that enhance the effectiveness of legal remedies. They operate in adversarial settings, handle technical disputes, and are positioned as a vital, accessible first tier of redress.

Yet, the introduction of Article 58a of the Statute of the Court of Justice in 2019 has complicated matters. This provision introduced a preliminary admission phase for appeals against decisions that have already undergone double review — first by a Board of Appeal, then by the General Court. Appeals are then only heard by the Court of Justice if they raise a question important to the unity, coherence, or development of EU law. The scope of this filtering mechanism extends to those cases originating in the Boards of Appeal of the EU agencies explicitly listed in Article 58a but also in any “independent board of appeal, set up after 1 May 2019 within any other office or agency of the Union”. Does this include AMLA’s ABoR? One could argue it does not, since the AMLA’s ABoR issues opinions which do not bind the AMLA. However, so does the Board of Appeal of the European Aviation Safety Agency which has been explicitly listed in Article 58a. Since the latter does not define what it means by ‘independent’, we cannot know for sure if the AMLA ABoR comes within its scope. In any event, given the administrative nature of the Boards of Appeal, the independence required of them under Article 58a of the Statute should not be set at the level of judicial independence. These questions evidently touch on the very architecture of EU remedies. In addition, the AMLA ABoR is competent to review sanctions that have a potentially “criminal” character. If the ABoR comes within the scope of the filtering mechanism, another question would then be whether there is and ought to be in compliance with Protocol 7 to the ECHR, which guarantees a right to (a judicial) appeal in criminal matters.

All these developments date from after the adoption of the Common Approach, showcasing how that framework has become even more outdated. It is needless to say that an updated Common Approach that properly recognizes the key role played by the Boards of Appeal in offering effective legal protection in areas of highly technical and scientifically complex decision-making would not be a luxury.

A Framework Outpaced by Practice

What emerges from all this is clear: the Common Approach is becoming obsolete. It does not reflect best practices, built up post-2012 each time new agencies were established or existing mandates were revised, anymore and neither does it provide institutional guidance on critical issues like seat selection, accountability, or legal remedies. More problematically, it currently coexists awkwardly with primary law developments, such as the introduction of Article 58a, that transform Boards of Appeal into bodies reminiscent of specialized courts — without granting them the judicial independence of specialized courts, and the Court’s confirmation that EU agencies are not EU institutions, meaning that it is up to the EU legislator and not the Member States to designate their seats.

In this regard, the AMLA Regulation illustrates how the Common Approach could be updated (reflecting best practices) and upgraded (turning it into a binding framework). It refines seat selection (even if imperfectly), revolutionizes accountability, and raises crucial questions about administrative review.

Ideally, such an update should do more than codify existing practice. It would articulate a principled vision for EU agencification: defining when and how agencies are created, what powers they wield, how and by whom they are held accountable, and how their decisions are reviewed. However, given that the proper place of EU agencies within the EU administration remains fundamentally contested between the political institutions, even a merely updated Common Approach, codifying existing best practices, would be a huge success.

Merijn Chamon is Professor of EU Law at the Vrije Universiteit Brussel and Visiting Professor at the College of Europe (Bruges)