The United States has accused Chinese artificial intelligence company Moonshot AI of using large-scale AI distillation techniques to replicate proprietary U.S. technology as it developed its K3 model.
Michael Kratsios, director of the White House Office of Science and Technology Policy, said the U.S. government has information indicating that Moonshot AI distilled Anthropic’s Fable model during the development of K3.
“We have information that Moonshot AI distilled Anthropic’s Fable for the development of its K3 model,” Kratsios wrote on X.
“To do this they developed a sophisticated internal platform to conduct large scale distillation against U.S. models, allowing them to quickly switch between multiple methods of access to avoid detection.”
Kratsios also alleged that Moonshot AI has obtained access to advanced Nvidia GB300-equipped servers, including servers in Thailand, potentially for training its AI models.
READ: Chinese CEO leaves US to launch successful AI startup despite ‘many opportunities’ (July 20, 2026)
“Moonshot AI has also acquired GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models,” he said.
The allegations highlight growing concerns in Washington over the methods Chinese AI companies use to develop increasingly capable models amid U.S. efforts to restrict China’s access to advanced AI chips and computing infrastructure.
Kratsios said the United States supports competition and innovation across the AI industry, including frontier models, specialized AI systems, open-source frameworks and open-weight models.
“The United States strongly supports the free and fair development of AI, including a thriving competitive ecosystem that spans frontier models, specialized systems, open-source frameworks, and open-weight models,” he said.
He distinguished between legitimate AI distillation, which can make AI models smaller and more efficient, and what he described as covert industrial-scale efforts to obtain proprietary technology.
“Legitimate AI distillation used to create smaller, more efficient models plays a vital role in this open innovation ecosystem,” Kratsios said. “However, large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable.”
Kratsios’s comments come as the Trump administration continues to focus on maintaining U.S. leadership in artificial intelligence while limiting the ability of foreign competitors to access advanced chips, computing power and cutting-edge AI technology.
The White House is also seeking to coordinate federal efforts around the use of AI in scientific research. Kratsios said the administration is working to bring together agencies involved in different areas of science and make them part of a broader national effort.
“The White House is helping bring together all of the different agencies that work on different domains to make them part of this larger effort on AI for science,” Kratsios said.
READ: Chinese AI models by Moonshot, Alibaba ramp up pressure on Silicon Valley (July 20, 2026)
The accusations against Moonshot AI underscore the widening technological competition between the United States and China. They also raise fresh questions about the use of model distillation, a technique that can have legitimate applications but has become a point of contention when used to reproduce the capabilities of proprietary AI systems.
Moonshot AI, the Beijing-based company behind the Kimi AI assistant, has emerged as one of China’s prominent AI startups. The company has drawn international attention for developing advanced AI models as Chinese firms compete with U.S. companies in the rapidly evolving generative AI market.
The allegations made by Kratsios have not been independently established in the statements provided, and it was not immediately clear whether Moonshot AI had responded to the claims.