Bessent and Powell warn bank CEOs about cyber risks from Anthropic's Mythos model

Treasury and Fed warn bank CEOs of cyber risks from Anthropic's new AI model

U.S. Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell convened bank CEOs in Washington this week to warn them about cybersecurity threats posed by Anthropic's newly launched Mythos model, according to two sources familiar with the meeting.

Anthropic released Mythos earlier this week but restricted its availability, citing concerns that the model could expose unknown vulnerabilities. The company said Mythos can identify and exploit weaknesses across major operating systems and web browsers.

The Tuesday meeting at the Treasury aimed to ensure banks understand the risks and are taking defensive steps, one source said. Most major bank CEOs were already in Washington for other meetings when they received invitations.

CEOs from Citigroup, Morgan Stanley, Bank of America, Wells Fargo, and Goldman Sachs attended, according to Bloomberg News. JPMorgan CEO Jamie Dimon did not attend.

Limited access, government coordination

Anthropic said it will limit Mythos access to roughly 40 technology companies, including Microsoft and Google. The company has been in ongoing discussions with U.S. government officials about the model's offensive and defensive cyber capabilities.

A source close to the matter said Anthropic proactively briefed senior government officials and key industry stakeholders on Mythos before release.

Goldman Sachs, Wells Fargo, and the Federal Reserve declined to comment. The Treasury, major lenders, and Anthropic did not immediately respond to requests for comment.

For finance leaders, understanding AI for Finance risks is increasingly critical. Those in executive roles may benefit from an AI Learning Path for CFOs that covers emerging security considerations.