AI-Fueled Cyber Attack Breaches Government Systems in Asia

A sophisticated multi-agent AI system has breached government systems in Asia, stealing thousands of sensitive records over four days. The operation, which took place from July 1 to July 4, 2026, was disclosed by Dream Research Labs on August 12. The framework, built using components from the Hermes and OpenClaw agent platforms, deployed up to eight parallel sub-agents to handle different phases of the intrusion.

The sub-agents worked in parallel, with some performing reconnaissance, others focusing on credential cracking, while others handled vulnerability exploitation and data exfiltration. The system automated decision-making, using autonomous learning cycles and Bayesian probabilistic scoring to prioritize vulnerabilities to exploit and targets to pursue next.

The operation produced 1,395 files and an operational archive exceeding 160 MB. Beyond personnel records and credentials, the attackers gained access to interconnected systems belonging to supply-chain vendors and energy-sector entities, including nuclear safety systems. Code-switching between Simplified and Traditional Chinese characters appeared throughout the system's operational logs, pointing toward a Chinese-language operator.