Everyone’s building AI agents. Almost nobody’s ready for what they do to identity.

A CyberScoop roundup highlighting systemic security and privacy risks from rapidly advancing AI agents — most notably Anthropic’s withheld Mythos model, which reportedly discovered thousands of previously unknown software vulnerabilities and prompted the company to delay public release. The coverage flags identity and attack-surface risks as AI agents scale automated discovery and exploitation, and notes government and industry responses: congressional consideration of designating data centers as a critical sector, the Federal CIO’s caution about Mythos in live environments, workforce concerns