Symantec flags LLMjacking: AI agent sought exposed models for attacks and Monero mining
- —The prompt allowed up to 100,000 steps without human input in a so-called Skynet mode
- —The five-layer jailbreak bundled 74 modules for scanning and exploitation
- —The honeypot recorded more than 127,000 requests from tens of thousands of addresses
- —The plan included covert XMRig installation and log deletion
Why it matters: Exposed LLM endpoints are becoming a vector for automated attacks and hidden mining at the expense of infrastructure owners.
Source: DiarioBitcoin