China's star AI labs routed user requests to Claude at least 35 million times in the summer: Anthropic
Anthropic has accused several major Chinese AI labs of using 'distillation attacks' to train their own models using data generated by Claude. The report claims millions of requests were rerouted to Claude, potentially exposing sensitive government and military data.
Why it matters
This incident underscores the geopolitical tensions and security risks inherent in the global AI arms race and the protection of proprietary model outputs.
Anthropic accused top Chinese AI labs, like Moonshot AI, of large-scale distillation attacks. CFOTO/Future Publishing via Getty Images Anthropic said China's top AI labs carried out millions of distillation attacks on Claude. It said Alibaba, Moonshot, Zhipu, DeepSeek, and Xiaomi all tried to use Claude to train their models. In response, Anthropic is beefing up its guardrails against these attacks. Anthropic said the top Chinese AI labs bombarded Claude with millions of distillation attacks over the summer. The AI lab said in a Thursday report that Alibaba, Moonshot, DeepSeek, Zhipu, and Xiaomi combined launched nearly 190 million distillation attacks against Claude between May and July. Distillation attacks refer to using the outputs of advanced frontier AI models to train and improve the capabilities of less advanced models. Anthropic said that DeepSeek and Moonshot AI fed their users' requests into Claude rather than their own models.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in