The European Banking Authority, together with the European Insurance and Occupational Pensions Authority and the European Securities and Markets Authority, has backed the European Systemic Risk Board warning that frontier artificial intelligence models create systemic cyber risks for the financial sector. The authorities said recent advances have made it possible to identify and exploit severe IT vulnerabilities much faster, raising the risk that AI-enabled cyberattacks could weaken the operational resilience of financial entities. They urged financial firms to adapt their cybersecurity capabilities and called on competent authorities to reflect these developments in supervision. The release says the existing European Union framework, including the Digital Operational Resilience Act and the AI Act, provides a basis for managing ICT and AI-related risks, but that the speed and scale of frontier models require stronger mitigation. The authorities pointed to earlier work under DORA, including their first annual report on major ICT-related incidents, in which they had already encouraged firms to reinforce cybersecurity measures as AI-driven tools evolve. In their role as overseers of critical ICT third-party providers, they are also engaging with those providers on the steps they are taking to manage the risks and maintain service continuity for the EU financial sector. The authorities will continue to monitor the development and use of highly cyber-capable frontier AI models and assess their impact on the financial sector. They are also working with national supervisors to clarify supervisory expectations and communicate them consistently to financial entities under the existing regulatory framework.

Original source Back to tracker