Pedro Machado, Member of the ECB Supervisory Board, addressed the role of artificial intelligence (AI) in banking supervision, highlighting both its potential and the ethical, operational, and supervisory risks. He noted that AI, like the internet before it, is a transformative technology that offers efficiency, improved analysis, and enhanced supervisory quality, but must be carefully guided to avoid errors, bias, or overreliance.
The ECB has implemented a structured approach, including the creation of a dedicated technology and innovation division, experimentation with AI use cases, and development of shared tools like Agora (data lake), Athena (document intelligence), and Virtual Lab (collaboration platform). AI tools such as Delphi, Medusa, Heimdall, and Navi assist supervisors by aggregating data, visualising relationships, detecting risks, and supporting complex analyses, while leaving human judgement central.
Machado stressed the ethical dimension of AI adoption, emphasizing that technology must be used responsibly, transparently, and in alignment with conduct principles. Key risks include AI inaccuracies, deskilling, explainability, cybersecurity, operational dependency, and AI-on-AI interactions. Compliance frameworks, operational risk assessments, and alignment with the EU AI Act are critical safeguards.
In conclusion, AI is a support tool, not a replacement for human supervision. By combining human judgement with machine assistance, supervisors can focus on high-value analyses, improve decision-making, and maintain financial stability in an increasingly complex and digital banking environment. Innovation, Machado reminded the audience, must be safe, ethical, and resilient.
