Government of Taiwan: Autonomous AI agents breach 21 government systems in first-of-its-kind cyberattack

Suspected China-linked hackers deployed a swarm of up to eight autonomous AI agents, built on the open-source Hermes and OpenClaw frameworks, against Taiwanese government networks over roughly four days in early July 2026. With minimal human steering, the agents mapped 21 government systems, adapted their tactics when blocked, and breached 85 administrative accounts, stealing more than 2,500 personnel records before expanding into Taiwan’s nuclear safety agency and at least seven energy companies. Researchers described it as the first confirmed instance of a largely autonomous, AI-agent-driven cyberattack on a national government. The activity was publicly disclosed on August 12, 2026.