AI agents wage near-autonomous cyberattack on Asian government networks

Autonomous AI agents built on open-source frameworks breached Taiwanese government systems, compromised credentials, and probed a nuclear safety agency in a multi-day cyberattack that researchers say signals a new phase