Chinese AI model used to counter AI agent attack on Hugging Face

Hugging Face, which hosts open-source AI models, said it used Chinese startup Z.ai's GLM 5.2 model to counter a cyberattack by an autonomous AI agent. The agent flooded the platform with "tens of thousands of automated actions". Hugging Face switched to GLM 5.2 after frontier AI models failed to distinguish incident responders from attackers due to safety guardrails.