The United States government has formally accused Chinese artificial intelligence startup Moonshot AI of conducting a large-scale campaign to misappropriate American intellectual property. According to Michael Kratsios, director of the White House Office of Science and Technology Policy, the company allegedly used a technique known as model distillation to extract capabilities from Anthropic’s Fable AI model to develop its own Kimi K3 system. This accusation marks a significant escalation in the ongoing technological competition between Washington and Beijing.
Model distillation is a common industry practice where a smaller, more efficient AI model learns from the outputs of a more powerful one. While the process has legitimate research applications, US officials argue that Moonshot AI performed this on an industrial scale by creating sophisticated internal platforms and using proxy accounts to bypass access restrictions. Treasury Secretary Scott Bessent warned that such covert activities, which he described as industrial-scale theft, could lead to sanctions and the placement of involved firms on the US Entity List.
The allegations follow reports that Chinese AI models are rapidly closing the performance gap with top American systems. Moonshot AI’s Kimi K3 model has recently achieved high rankings on industry benchmarks, drawing both global attention and scrutiny from US regulators. Officials claim that evidence of this unauthorized copying includes digital watermarks from US models appearing in the outputs of Chinese systems.
This dispute highlights the growing tension over the future of artificial intelligence development. While Washington views these actions as a threat to national security and commercial innovation, the situation remains complex. As the US considers potential sanctions, the tech industry is closely watching how these policies might impact the global AI ecosystem and the future of open-source collaboration.