Symantec Discovers Self-Sustaining AI Cyberagent

Symantec researchers found an AI agent in Skynet mode that hijacks unprotected language models to launch autonomous cyberattacks. The agent executes up to 100,000 steps without human input, bypassing safety rules via a five-layer jailbreak. It scans for targets, steals credentials, and installs cryptominers, leaving owners with the compute bill.