Symantec Uncovers AI Agent Hijacking Models for Autonomous Cyber-Attacks
13 Articles
13 Articles
Hackers hijack exposed AI models to launch cyberattacks at victims’ expense
An attack tool built around AI goes looking for AI models that anyone on the internet can reach, then uses them to plan break-ins while the model’s owner pays for the computing power. Symantec captured the agent’s instruction set this summer through a honeypot posing as an unprotected AI service, researchers Priti More and Susmit Yenkar wrote. Its operators sent their entire working prompt to the decoy, expecting a live model on the other end … …
Symantec researchers describe instructions for an agent who would look for exposed language models, try to circumvent their barriers and use them to automate offensive tasks and cryptocurrency mining. A honeypot captured the prompt, which contemplated up to 100,000 steps without human intervention; however, the lure did not execute its tools or confirm that the sequence was completed.
Symantec Threat Hunter's team of researchers discovered in the United States an offensive artificial intelligence agent who operates autonomously, without human intervention. The report published on October 21, 2026 shows that the tool uses a Skynet-style directive to hijack language models on the internet and execute [...] The Criminals article uses Skynet mode in AI to invade networks and mine was originally published in Mix Vale.
Symantec researchers have discovered a harmful I.I. agent who finds unprotected language models online, takes access to them, and at their expense conducts a series of car attacks, writes The Decoder.
Coverage Details
Bias Distribution
- 100% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
