Vapor Panda used ARTEX AI agents against more than 200 organisations, CrowdStrike says

CrowdStrike says a China-nexus group it tracks as Vapor Panda used ARTEX, an open-source multi-agent AI penetration-testing framework, in attacks on more than 200 organisations across 14 countries. BleepingComputer reported the findings on November 2.

The operation began on October 17, according to the report. Targets included government ministries, telecoms providers, and firms in pharmaceuticals, semiconductors, defence, nuclear and finance. The actor compromised at least nine organisations and ran password spraying and credential scanning against at least 12 other high-value targets.

Confirmed victims named in the report include telecoms providers in the United States and Belgium, government organisations in Greece, the Middle East and Latin America, and US pharmaceutical firms. In the Greek case, the agents forged an administrator token and read mailboxes using authentication keys from a 2022 mail-server directory export.

The agents did reconnaissance, exploited vulnerabilities, deployed web shells and grabbed credentials, with operators issuing instructions in Chinese and stepping in at decision points. CrowdStrike says they were told to skip Windows hosts protected by Falcon. In one intrusion the agents are believed to have consumed more than 1.4 billion input tokens.

Separately, the Dutch Institute for Vulnerability Disclosure says ARTEX was the tool behind its September breach, which went from initial access to root in 12 minutes by chaining two zero-days. DIVD logged over 450,000 requests.

What defenders can act on is thin: the report gives no indicators, and the framework's creators have since made it proprietary. The Falcon-skipping detail is the one concrete signal, and it's a reason not to treat endpoint coverage as the whole picture. We'd want the CrowdStrike write-up before drawing more from it.