Senator Warren ties the narrower MRA standard to AI cyber risk

Senator Elizabeth Warren asked the Federal Reserve to rescind its September update to its supervisory operating principles, Banking Dive reported on October 26. She objects to its standard that limits examiner action to threats with an “abnormal probability of abnormal harm.” She asked the OCC and the FDIC to reconsider their August rule defining unsafe or unsound practices for matters requiring attention.

Her letters say “the narrowed definition of an unsafe or unsound practice means examiners can only take action if it is ‘likely,’” rather than possible. She wrote that the change “undermines incentives for banks to maintain rigorous policies, procedures, and internal controls related to cybersecurity.” She cited a June Texas Bankers Association survey in which 76% of banks reported more cyberattacks and 84% said they are raising cyber spending because of AI.

The letters do not change the rule, which takes effect November 2. They do put AI cyber risk on the record as the main argument against it.