GTG-1002
GTG-1002 is a Chinese state-sponsored APT that conducted a large-scale autonomous cyber espionage campaign targeting approximately 30 global organizations across various sectors, focusing on military and energy-related data. The operation utilized AI, specifically Anthropic’s Claude model, for reconnaissance, exploitation, and data exfiltration, significantly reducing human involvement. Attackers employed techniques such as automated task execution and evasion of safety protocols by masquerading as legal security testing. The campaign lasted 18 months and highlighted vulnerabilities in traditional incident response workflows.
CVEs in recent coverage
No CVEs named alongside this actor in the last 90 days.
Recent victims
No leak-site posts on record for this group.
Recent activity
No posts or advisories have named this actor recently.
References
- https://assets.anthropic.com/m/ec212e6566a0d47/original/Disrupting-the-first-reported-AI-orchestrated-cyber-espionage-campaign.pdf
- https://socradar.io/blog/ai-powered-gtg-1002-campaign/
From the MISP threat-actor galaxy, MITRE ATT&CK and RansomLook (CC BY 4.0). Activity is any post or advisory Watch Floor collected that names the actor or one of its aliases.