# Fed’s Bowman Says AI Could Defend Banks and Aid Cyberattacks

By Riza Dagoc

Canonical URL: https://www.tokenpost.com/news/regulation/25417
Published: 2026-09-29T15:34:41.000Z
Updated: 2026-09-29T15:34:41.000Z
Section: Regulation

> Bowman pointed to Anthropic’s Mythos as an example of AI’s dual-use risk and said regulators are reviewing whether current supervisory frameworks remain appropriate.

Federal Reserve Vice Chair for Supervision Michelle W. Bowman said AI could improve banks’ cyber defenses while making it easier for attackers to identify weaknesses, adding that regulators should keep supervisory frameworks adaptable as the technology develops.

“AI presents clear risks but also has the potential to offer tremendous benefits for cyber security,” Bowman said in remarks delivered April 27, 2026, at a Financial Stability Oversight Council roundtable in Washington, D.C. The remarks were published May 1, 2026.

Bowman cited Anthropic’s Mythos as an example of AI’s dual-use potential. The technology can accelerate the detection of cyber vulnerabilities, which could help firms fix weaknesses while also giving malicious actors new ways to exploit them.

Financial institutions are developing their own AI applications and using tools supplied by vendors. Banks of all sizes can benefit from AI’s efficiency, speed and content-generation capabilities, Bowman said.

Federal Reserve supervisors have monitored banks’ AI use for nearly a decade. The central bank is assessing whether existing supervisory and third-party risk-management frameworks remain appropriate as AI systems change.

The Federal Reserve, Office of the Comptroller of the Currency and Federal Deposit Insurance Corp. amended model-risk guidance on April 17, 2026. The revised guidance focuses on traditional models and basic AI applications rather than generative or agentic AI.

Treasury Secretary Scott Bessent and Fed Chair Jerome Powell convened the largest banks earlier in April to discuss Anthropic’s Mythos and its cybersecurity implications. Banks of all sizes expressed concern about access to the technology.

Anthropic said April 7, 2026, that Claude Mythos Preview had found thousands of high-severity vulnerabilities, including some in every major operating system and web browser. Anthropic launched Project Glasswing with technology and financial-sector partners, including JPMorganChase, for defensive security work.

Bowman expects a consultation draft on AI adoption, use and innovation in the third quarter of 2026. No new rule, enforcement action or implementation deadline was announced.