GTIG Says Customized PentAGI Ran Recon Against Asian Targets
Summary
Google Threat Intelligence Group and Mandiant report that threat actors now deploy autonomous multi-agent frameworks such as CAIRN and PentAGI in live intrusions, not just manual prompting. GTIG says a customized PentAGI platform ran reconnaissance and scanning against Asian organizations from late August to early October 2026. In LLMJacking cases, Mandiant says single-day fraudulent cloud charges exceeded $100,000.
Key Points
- GTIG reports CAIRN, an open source OODA-loop framework, appears in intrusions targeting US aerospace organizations.
- CyberStrikeAI pairs commercial and open-weight LLMs with more than 100 security tools through a Go-based backend that doubles as a command-and-control server.
- GTIG finds an exposed server running exploit_agent, a framework that drafts and executes exploits without human input against French media and education sites.