AI incidents at banks: governance finding

How many AI incidents at banks since 2022 were cases where a court, regulator, auditor or the institution itself found a problem in how AI was governed, marketed or used (bias, litigation, regulatory findings, AI-washing)?

As of Oct 21, 2026, the register holds 24 publicly reported AI incidents at banks and fintechs where a court, regulator, auditor or the institution itself found a problem in how AI was governed, marketed or used (bias, litigation, regulatory findings, AI-washing) since January 2022. 11 of the 24 records involve systems the institution or a regulator described as AI; 7 are automated systems not described as AI, and 6 are machine-learning systems not labelled AI on the record. Countries: United States (17), United Kingdom (2), Germany (1). Average severity 2.3 on a 1–5 scale. Status: 2 confirmed by the institution, 7 regulator finding, 8 court filing, 5 settled, 2 disputed. The most recent is "U.S. Bank customers sue Pegasystems alleging its Voice AI secretly analysed their customer-service calls" (Oct 2, 2026).