Chinese-Speaking Threat Actor Uses DeepSeek AI for Autonomous Cyberattacks on Vulnerable Servers

A Chinese-speaking threat actor is utilizing the DeepSeek AI model alongside the open-source Hermes Agent to execute autonomous cyberattacks on exposed servers with minimal human oversight. The attacks target vulnerable systems, though specific vulnerabilities, affected software versions, or CVE identifiers were not disclosed. The campaign leverages AI-driven automation to identify and exploit weaknesses without requiring continuous manual input. No exact timeline, geographic focus, or quantified impact (e.g., number of compromised systems) was provided in the report. The technique demonstrates an evolution in threat actor tactics by integrating commercial AI tools for offensive operations.