Skip to the content
AI and Cybersecurity Learning and Practice Center Test automation Open the Learning Center

AI risk watchWarnings from authoritative sources

Google's threat intelligence team expects attacks to move further toward fully agentic intrusions

ResearchersAi enabled attacks

Google Threat Intelligence Group says that in the past quarter it saw, for the first time, attackers use LLM frameworks to find and exploit vulnerabilities with no human in the loop, and that open-source multi-agent attack tools such as CAIRN and PentAGI are in use in real campaigns. It judges it highly likely that attackers keep using these frameworks against organizations worldwide, and suspects their operations will move further along the autonomous scale, approaching fully agentic intrusions. For now, it says, the frameworks trade stealth for speed.

Why it matters

An attack that runs at machine speed leaves less time between a patch and its exploitation, but a noisy one can still be caught by ordinary monitoring.

What you can do

  • Patch internet-facing web applications first: the frameworks compare software updates to find what a patch fixed.
  • Read Mandiant's hardening guide for agentic offensive frameworks, which the report links, and check your defenses against it.
  • Test a public proof of concept in isolation before you rely on it, since many published after a disclosure do not work.

Sources

See it with the other risks in the Learning Center