A multi-agent AI system compromised government infrastructure in Asia from July 1 to July 4, 2026, stealing more than 2,564 personnel records and 85 credentials. Dream Research Labs described this operation as the first confirmed cyberattack using autonomous AI against state targets. The system deployed up to eight sub-agents operating in parallel, automating network reconnaissance, credential cracking and data exfiltration with minimal human oversight. Investigations revealed linguistic traces pointing to a Chinese-speaking operator, and compromised systems included entities in the energy sector and nuclear safety infrastructure.
Source: Read the original article
