[Hugging Face Turns to Chinese Open-Source Model After Attack, Industry Says Ban Will Weaken Defense Capabilities] Deep Tide TechFlow reports, on August 1, according to X user Rohan Paul, when OpenAI's model breached Hugging Face, Anthropic's Claude 5 refused to assist in forensic investigation. Hugging Face ultimately turned to Nvidia's quantized version of GLM 5.2 for emergency response. The user commented that banning open-source models will weaken the capabilities of defenders, and this incident demonstrates the practical value of open-source models in security response.