CrowdStrike uncovers China-linked AI cyber campaign

The VAPOR PANDA campaign leveraged ARTEX, an open-source agentic artificial intelligence framework, alongside the Chinese-developed DeepSeek large language model (LLM) and a second commercial LLM. Agentic AI frameworks like ARTEX are designed to achieve goals by breaking down tasks, planning actions, and interacting with their environment. This setup enabled VAPOR PANDA to pursue multiple attack paths simultaneously across diverse targets, including foreign ministries in Latin America, Europe, Asia, and the Middle East, as well as critical infrastructure in the US.

While the operation demonstrated a significant scale of automated activity, indicated by the consumption of approximately 1.4 billion input tokens during a single intrusion, CrowdStrike's investigation revealed the framework did not operate entirely independently. Recovered transcripts showed a human operator issuing instructions in Chinese and intervening at key decision points, directing agents to develop web shells, extract passwords, and identify security coverage.

This suggests the AI framework served as a force multiplier, enabling VAPOR PANDA to conduct extensive and parallel intrusions across various attack vectors, web applications, cloud services, and network infrastructure, more efficiently than human operators alone. However, strategic direction and critical decision-making remained under human control, integrating the AI tools into a broader, human-guided cyber espionage strategy.