Google warns ‘the era of autonomous attacks has arrived’ as hackers deploy AI exploit tools
Google Threat Intelligence Group says attackers are now using autonomous LLM frameworks that scan software updates, identify patched flaws and build exploits with little or no human involvement.

Google Threat Intelligence Group (GTIG) said on November 16 in its latest AI Threat Tracker report that “the era of autonomous attacks has arrived.” According to the report, attackers are using autonomous frameworks built on large language models to continuously scan software updates, work out which flaw was patched and build an attack for it, all without human involvement.
Tools in the wild
Researchers found an exposed server hosting an AI-powered hacking system called exploit_agent, Newsweek reported. It was designed to examine developers' updates, identify the fixed weaknesses and create attacks targeting machines that had not yet been patched. The server also held data stolen from victims along with tools for searching compromised networks, breaking into accounts and installing malware; evidence suggested the operator was interested in French media, entertainment and education websites.
Another framework, CAIRN, was observed in intrusion activity that included targets in the US aerospace sector, while a tool based on PentAGI was used to scan organizations across Asia. The report also described criminals stealing and reselling access to paid AI services and hijacking companies' cloud accounts to run AI workloads; some victims faced more than $100,000 in fraudulent charges in a single day, with cumulative costs reaching $1 million in the most serious cases.
Defense at machine speed
“For the most part, attackers aren't using new groundbreaking methods to break into systems—they are still going after known, unfixed issues, just with much greater speed and scale,” said John Hultquist, chief analyst at GTIG. Google said autonomous tools still leave recognizable traces and some AI-generated exploits fail, but it expects the technology to improve. Arguing that machine-speed threats cannot be defended at human speed, the company pointed to Google AI Threat Defense, which combines Gemini and other frontier models with Wiz, CodeMender automated patching and Mandiant expertise.
Get the latest news delivered to your inbox.
Related articles





