The European AI Act’s Impact on Financial Markets: From Governance to Co-Regulation

The study explores the general regulatory challenges arising from the need to mediate between a horizontal approach to AI regulation and the sector-specific dimensions of financial markets, with regards to the sectoral regulations that may already regulate some AI tools in the financial sector, the structural features of AI-driven financial services, as well as the traditional objectives of financial regulation. In this vein, this paper intends to identify the gaps left open by the AI Act’s requirements for a full control over financial AI-related risks. In second stance, it aims at providing first guidelines for the implementation of the AI Act’s horizontal requirements in the financial sector. To these ends, the analysis demonstrates the relevance of the governance strategies regarding ICT risks that financial institutions – and their boards of directors – will have to develop under the DORA, as an effective venue for addressing detected regulatory gaps in the field of financial AI. Outside the scope of single financial institutions’ compliance, and shifting to a market-wide perspective, the acknowledgment of the shortcomings related to the tools of supervised co-regulation (as regulatory sandboxes) proposed by the AI Act, confirms the importance of the enactment of sound supervisory policies by financial authorities. In this respect, it is shown how the AI Act’s provisions on AI supervision, matched with the rules under the DORA, pose fruitful legal grounds for the consolidation of financial AI supervisory schemes, based on the collaboration between different financial authorities, and between financial and other technology-relevant authorities, as the European Artificial Intelligence Board.