Chinese Hackers Deploy AI Agents in Automated Attacks on 200 Organizations
Chinese hackers are using ARTEX, an open-source multi-agent AI penetration-testing framework, to attack more than 200 organizations across 14 countries in a campaign that began on October 17. The targets include government ministries, telecommunications providers, and organizations in the pharmaceutical, semiconductor, defense, nuclear, and financial sectors. At least nine organizations are compromised, while password spraying and credential scanning hit at least 12 other high-value targets.
CrowdStrike researchers track the threat actor as Vapor Panda and describe the operation as an escalation in China-nexus use of AI agents, moving from single intrusions to multiple simultaneous engagements directed by a human operator. Confirmed victims include telecom providers in the United States and Belgium, government organizations in Greece, the Middle East, and Latin America, and U.S. pharmaceutical firms. In an attack on a Greek government entity, the AI agents forge an administrator token and read mailboxes using authentication keys from a 2022 mail-server directory export.
The operators issue instructions in Chinese and intervene at critical decision points while the AI agents perform reconnaissance, exploit vulnerabilities, deploy web shells, and obtain credentials. The attackers explicitly instruct agents to skip Windows machines protected by CrowdStrike Falcon and focus on unprotected hosts, consuming more than 1.4 billion input tokens in one intrusion. Vapor Panda also uses DeepSeek and a second unidentified commercial language model to coordinate agents, and deploys the BLOODALCHEMY backdoor against government entities in Latin America.