Symantec finds AI agent that hijacks unprotected AI models for attacks
Symantec has found a malicious AI agent that takes over unprotected AI models online. It uses them to scan for targets, steal logins and databases, and install cryptominers. It can run up to 100,000 steps with no human help. The models' owners pay the computing bill. Symantec caught the plan with a decoy system.
Where this came from. One outlet reported this. We read it, then wrote the sixty words above.