Chinese-linked group used AI penetration framework ARTEX against more than 200 organizations
Confirmed post-exploitation at nine organizations shows AI-driven tooling moving from testing into live intrusions.
CrowdStrike attributes a campaign using the Go-based, AI-driven penetration testing framework ARTEX to a China-based intrusion group it tracks as Vapor Panda, which overlaps with Trend Micro's Earth Estries cluster. Since October 17, 2026, the group reportedly targeted more than 200 organizations worldwide, including foreign ministries, a Middle East interior ministry, U.S. semiconductor, defense and pharmaceutical firms, and telecom providers. Post-exploitation was confirmed at nine organizations, where the attackers deployed BLOODALCHEMY, a ShadowPad derivative, and at least 12 more faced password spraying and credential scanning. The operation reportedly pairs ARTEX with DeepSeek and an unnamed subscription LLM, using planner and worker agents overseen by a human operator, with planning agents alone consuming about 1.4 billion input tokens in one intrusion. The disclosure coincides with the developer Li Puhua taking ARTEX closed source after its rapid uptake among threat actors, and with four newly documented China-linked groups, including Overcast Panda, which uses the FlowCloud backdoor.
Why it matters
The campaign shows how an AI-driven penetration testing framework can be run by a human operator with planner and worker agents to scale intrusions across many targets at once. Confirmed breaches at nine organizations, with password spraying and credential scanning against at least 12 more, show the approach is already in active use rather than hypothetical. The closed-sourcing of ARTEX after its wide uptake among threat actors means the tool's future spread is now harder to track from outside, which leaves defenders with less visibility into how it is being adapted.
Matters or hype?
Does this story matter, or is it hype? Vote before you see what everyone else said.