Anthropic Accuses Chinese Firms of Model Distillation

Anthropic has publicly accused three Chinese companies—DeepSeek, Minimax, and Moonshot AI—of distilling their Claude model, which involves extracting knowledge from a proprietary AI model to create a new one. The controversy highlights ongoing concerns regarding intellectual property and the security of AI models. Anthropic's blog discusses methods for detecting and preventing such distillation attacks, emphasizing the importance of safeguarding AI technologies. This situation underscores the challenges faced by AI developers in protecting their innovations in a rapidly evolving landscape.