Particle.news
Download on the App Store

BaFin Expands Oversight of AI in Banks and Insurers

The regulator will begin sampling financial firms’ AI systems now and will impose stricter rules on credit-scoring and insurance pricing designated as high-risk from December 2027.

The logo of Germany's Federal Financial Supervisory Authority BaFin (Bundesanstalt fuer Finanzdienstleistungsaufsicht) is pictured outside of an office building of the BaFin in Bonn, Germany, April 15, 2019. REUTERS/Wolfgang Rattay

Overview

  • BaFin announced it will immediately start monitoring a sample of AI applications used by banks, insurers and other authorised financial firms to check compliance with new rules.
  • The move uses powers granted under the EU AI Act and lets BaFin enforce transparency rules, block prohibited AI practices and levy fines to protect fundamental rights.
  • BaFin has identified creditworthiness models and insurance risk‑pricing systems as 'high-risk' AI and will begin targeted high‑risk oversight in December 2027.
  • The regulator will assess whether firms meet transparency requirements for customer-facing tools, avoid biased use of sensitive personal data and train staff on AI literacy.
  • Firms face trade-offs between faster loan approvals and claims processing from AI and new compliance demands that will require stronger model documentation, governance and staff upskilling.