White House Accuses China's Moonshot of Using Restricted Nvidia Chips and Distilling Anthropic's Fable for Kimi K3

On July 22, 2026, Michael Kratsios, director of the White House Office of Science and Technology Policy, accused Chinese AI company Moonshot of illegally obtaining restricted Nvidia GB300 processors in Thailand and distilling Anthropic's Fable AI model to create its Kimi K3 system, escalating the US-China AI rivalry into a major enforcement and sanctions crisis. For broader context, explore our AI News.

Allegations of Restricted Chip Access

Michael Kratsios stated that Moonshot reportedly gained access to Nvidia's high-end GB300 processors within Thailand. These specific chips are subject to stringent US export controls, designed to limit advanced technology transfer to certain foreign entities. The accusation points to a potential circumvention of these controls, raising concerns about the effectiveness of current export regulations in preventing the acquisition of advanced AI hardware by targeted companies.

Accusations of Model Distillation

In addition to hardware concerns, Kratsios also alleged that Moonshot distilled Anthropic's Fable AI model. Distillation typically involves training a smaller, more efficient model to replicate the performance of a larger, more complex 'teacher' model. If confirmed, this would imply that Kimi K3, Moonshot's 2.8-trillion-parameter open-source system, leveraged intellectual property from a leading US-developed AI model. Moonshot claims Kimi K3 rivals prominent US frontier models, a claim that takes on new context given the distillation allegations.

Kimi K3 and the Broader AI Landscape

The accusations against Moonshot follow closely on the heels of the July 18 release of Kimi K3 and Alibaba's Qwen3.8 preview. Both Chinese models were presented as rivals to established US frontier models. Moonshot has also committed to releasing the full model weights for Kimi K3 on July 27. This rapid development and release of advanced AI models from China, coupled with the White House's allegations, underscore the intense competition and geopolitical tensions surrounding AI development.

Implications for US-China AI Rivalry

The White House's allegations mark a significant escalation in the US-China AI rivalry. What was previously a competition in technological advancement is now moving into an arena of enforcement and potential sanctions. The situation is also reportedly creating internal debate among Trump's AI advisors, with discussions centering on whether to prioritize open-source AI competition or to implement more stringent government controls to protect US technological advantages. This internal division highlights the complex policy challenges faced by governments attempting to regulate rapidly evolving AI technologies in a globalized environment.

Conclusion

The White House's accusations against Moonshot regarding the use of restricted Nvidia chips and the alleged distillation of Anthropic's Fable model for Kimi K3 represent a critical juncture in the US-China AI competition. These claims, made on July 22, 2026, elevate the rivalry to an enforcement and sanctions crisis, with potential long-term implications for global AI development and trade policies. The coming weeks may reveal further details as the international community observes responses from both Moonshot and the US government.

Sources

• https://www.theverge.com/ai-artificial-intelligence/969454/chinas-moonshot-trained-its-ai-model-on-a-restricted-nvidia-chip-white-house-claims
• https://www.theverge.com/ai-artificial-intelligence/967781/chinese-ai-models-open-source-moonshot-kimi-k3-alibaba-qwen
• https://www.technologyreview.com/2026/07/20/1140675/chinas-ai-models-have-trumps-ai-world-at-war-with-itself/
• Artificial Intelligence
• Emma Roth Profile and Activity | The Verge