Moonshot AI Accused of Stealing from OpenAI and Anthropic Models
US-based OpenAI has disrupted a large-scale effort by users linked to Chinese AI lab Moonshot AI to extract protected reasoning from its models. The company claims that it caught the attempt, which was tied to users affiliated with Moonshot AI.
The targeted reasoning is the step-by-step problem-solving process that makes modern AI models useful for complex tasks. This kind of extraction is known as 'distillation,' where a powerful model's outputs are fed into a competing system to train it.
OpenAI is not the only American lab pointing fingers at Moonshot. Anthropic has accused the company of sending nearly 300,000 requests through fraudulent accounts for distillation purposes. Michael Kratsios, director of the White House Office of Science and Technology Policy (OSTP), has also accused Moonshot of stealing from Anthropic's Fable model.
Moonshot denies the allegations, claiming that its results come from legitimate innovation, not from siphoning American models. The company released its flagship Kimi K3 model on July 16, 2026, which has 2.8 trillion parameters and is open-weight, allowing outside developers to download and run it themselves.