Anthropic Leak Exposes “Claude Mythos,” Sending Cybersecurity Stocks Into Panic
On March 27, a technical misconfiguration inside Anthropic’s content management system accidentally released nearly 3,000 confidential files to the public. Buried within the leak were detailed documents outlining Claude Mythos, described internally as the company’s most powerful AI model to date.
XingChi

