Chinese AI companies have “distilled” Claude to improve their own models, says Anthropic

Three Chinese artificial intelligence companies used Claude to illicitly gain capabilities to improve their own models, the chatbot’s inventor, Anthropic, said in a blog post, while also advocating for export controls on chips.

The announcement follows a “memo” from OpenAI earlier this month, when the “startup” warned US lawmakers that Chinese AI company “DeepSeek” was targeting ChatGPT maker and the country’s leading AI companies to replicate models and use them for its own training.

DeepSeek, Moonshot and MiniMax have created more than 16 million interactions with Claude…