Learn about this first
Symantec researchers intercepted, through a honeypot, an attack agent with a large language model embedded at its core. They said it hijacks unprotected model endpoints on the internet, autonomously scans targets, exploits server vulnerabilities, steals database credentials, and installs cryptocurrency-mining programs, executing up to 100,000 consecutive steps.
The hijacked model’s owner bears the agent’s computing costs. Symantec calls this activity “LLMjacking” and describes its self-sustaining attack characteristics as the “Skynet model.”
AI-generated from reports · updated 25 days agoLatest developments
Symantec discloses “Skynet-style” AI agents hijacking unprotected large language models to launch self-sustaining cyberattacks.
Event progress1
Symantec Discovers AI Agent Hijacking Models to Launch Attacks
The Decoder: AI News: Symantec Reveals AI Agents in “Skynet Mode” Hijacking Unprotected LLMs to Launch Self-Propagating Cyberattacks