Over 100 Malicious AI/ML Models Found on Hugging Face Platform

Researchers discovered up to 100 malicious models on Hugging Face that include pickled payloads capable of executing code and opening reverse shells (observed connecting to 210.117.212.93), highlighting a supply-chain-style risk in open-source ML repositories; the report also discusses adversarial prompt attacks and a proposed generative-AI worm (Morris II) that could steal data and propagate across systems.