FDIC flags cyber risk, is testing AI tools and may issue RFI on bank AI use

Reporters asked whether regulators are keeping pace with banks’ use of artificial intelligence and whether cyber risk remains the top threat to financial institutions.

Chairman Travis Hill said cyber risk remains a focal supervisory concern and that the FDIC is rethinking how IT exams should be conducted to emphasize protections against cyber intrusions rather than checklist exercises. “We’re all thinking hard about what the future of IT exams should look like,” he said.

On AI specifically, Hill said the agency has been working on AI matters for six to nine months and recently unveiled Microsoft Copilot to the entire agency as an internal productivity tool. He said the FDIC is still in early stages of integrating AI into its operations but sees potential to increase productivity. Hill also said the FDIC and its peer agencies have discussed issuing a Request for Information to better understand banks’ AI use and noted ongoing coordination about how to supervise AI use at institutions.

Hill mentioned government interest in the Mythos system and said agencies are focused on what it means for protecting sensitive information. He did not provide details on any new AI guidance or supervisory requirements during the briefing.