Anthropic CEO Dario Amodei warned JPMorgan Chase CEO Jamie Dimon on Tuesday of a critical six to 12-month window to address tens of thousands of software vulnerabilities uncovered by the company's Mythos AI model. Amodei described the situation as a potential surge in breaches and ransomware attacks on hospitals, schools, and banks as AI identifies flaws faster than organizations can patch them.
Early testing with Mozilla showed Mythos identifying 271 vulnerabilities in Firefox in a single pass, demonstrating AI's ability to scan vast codebases far quicker than human researchers. Anthropic has restricted the model to a small group of partners under Project Glasswing to fix vulnerabilities before similar tools become widely available. Critics, including OpenAI CEO Sam Altman, have dismissed the warnings as fear-based marketing aimed at justifying restricted access. Despite public feuds, the U.S. government is reportedly using Mythos to scan classified networks for vulnerabilities.