Open-Source AI Attack Tool ARTEX Now Used by Second China-Linked Group Against 200+ Targets

A China-linked hacking group has used an open-source AI attack framework to target more than 200 organisations worldwide in less than a fortnight, according to CrowdStrike. The group, tracked as VAPOR PANDA, began using the ARTEX framework on October 17. Its targets included foreign ministries, US semiconductor, defence and think tank organisations, and telecommunications providers. CrowdStrike confirmed follow-on activity at nine organisations, where attackers deployed a backdoor and carried out credential spraying, network scanning and reconnaissance.

ARTEX is not an AI model but a harness: software wrapped around a large language model that gives it tools such as a command shell, a shared record of discoveries, and a planner that splits an intrusion into tasks for multiple agents. It can connect to models from Anthropic and OpenAI, or any service using OpenAI's programming interface. VAPOR PANDA mainly ran it on DeepSeek, with a human operator issuing instructions in Mandarin. The developer withdrew the original code in October because of misuse, Google Threat Intelligence Group and Mandiant said, but copies still circulate, including a commercial edition called ARTEX PRO.

This is the third major ARTEX campaign disclosed this month. Microsoft linked it to Storm-0940, which used it in the ConnectWise Cloud Backup compromise and in several cases moved from initial targeting to privileged cloud access in under three hours. CrowdStrike also found it used against South Korean financial institutions, likely by an individual based in China.