Hugging Face uses open-weights Z.ai GLM 5.2 to battle attacker after commercial frontier model refusal

Hugging Face Inc., an open-source artificial intelligence platform often described as the “GitHub of machine learning,” found itself forced to use an open-weights model to respond to an agentic AI attack after the safety guardrails on commercial AI models blocked requests. Last week, Hugging Face said it detected a breach from an attacker using an […] The post Hugging Face uses open-weights Z.ai GLM 5.2 to battle attacker after commercial frontier model refusal appeared first on SiliconANGLE .