Anthropic's Mythos AI Cracked 'Almost All' US Classified Systems in Hours, NSA Chief Reportedly Said

Anthropic's Mythos AI Cracked 'Almost All' US Classified Systems in Hours, NSA Chief Reportedly Said highlights a reported security breach claim involving Anthropic’s most powerful model. Senator Mark Warner said on June 11 that General Joshua Rudd, head of both the NSA and the Pentagon’s Cyber Command, told him directly that Anthropic’s Mythos “broke into almost all” US classified systems during testing, not in weeks but in hours. The allegation has not been publicly confirmed by any government agency, and neither the scope nor the specific systems affected were disclosed. Alongside the warning, the Trump administration reportedly blocked foreign access to Anthropic’s two most advanced models, Mythos and Fable, revoking permissions for government agencies, banks, and major firms, including within the Five Eyes partners. Britain’s AI Security Institute was also locked out. Experts cited by the report questioned the ban’s effectiveness and noted it may be difficult to sustain, while Anthropic did not respond to a request for comment.