The Fed rang the alarm about Anthropic's Mythos AI model — but had to go months without it
The central bank didn't have access to Claude Mythos Preview as of mid-July, even as other institutions were racing to patch their vulnerabilities.
AI Summary
The Federal Reserve and Treasury Department warned top banks in April about Anthropic's Claude Mythos Preview AI model, which excelled at identifying software vulnerabilities, but the Fed itself lacked access to the model for at least three months while other institutions began patching weaknesses. The central bank was still trying to obtain Mythos as of July 15, leaving it potentially exposed to the very threats it had flagged.