U.S. Agencies Accuse China AI Firms of Distilling Claude, GPT, Gemini, and Grok

US cybersecurity and intelligence agencies have accused China-based artificial intelligence companies of extracting proprietary functionalities from American AI models through distillation attacks. This activity is said to be happening on a large scale and is a key part of their AI development strategy. The targeted models include Claude, GPT, Gemini, and Grok. The accusations suggest that these Chinese companies are using the extracted information to advance their own AI capabilities. The distillation attacks are described as a systematic effort to obtain sensitive information from these American models. The scale of the activity is characterized as industrial-scale, indicating a significant and coordinated effort.