Taiwan confirms AI-assisted attack on government agencies

Taiwan's Ministry of Digital Affairs officially confirmed on August 13, 2026 that monitoring teams detected foreign activity combining human operation and AI-agent assistance in an incident found in July; the affected agencies completed remediation, and no country was attributed.

Israeli cyber company Dream described a multi-agent intrusion affecting Taiwanese agencies but did not provide underlying data to reporters.

Even with a human operator in control, agentic systems can accelerate reconnaissance and credential theft. Identity controls, egress monitoring, tool telemetry and fast credential rotation are essential. There is no evidence to call this fully autonomous hacking or to attribute it to China.