The White House has claimed that China’s AI startup Moonshot AI has secretly leveraged Anthropic’s own technology during the development of its own AI model K3. The allegations came from the White House Office of Science and Technology Policy Director Michael Kratsios in a write-up on X.
White House Slams Moonshot AI For Allegedly Copying Anthropic’s Fable Model
The U.S. government has evidence that Moonshot AI’s K3 model development was facilitated by Anthropic’s Fable models, Kratsios said. He also claimed that the company has developed a covert system to carry out the distillation of AI models in bulk without being noticed.
“We have information that Moonshot AI distilled Anthropic’s Fable for the development of its K3 model,” Kratsios wrote.
The company also developed a “a sophisticated internal platform to conduct large scale distillation against U.S. models,” he alleged. This enabled it to “quickly switch between multiple methods of access to avoid detection,” the White House official added.
We have information that Moonshot AI distilled Anthropic’s Fable for the development of its K3 model.
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…
— Director Michael Kratsios (@mkratsios47) July 22, 2026
During the process, Kratsios says that Moonshot AI also increased its computing capacity. He claimed that it had acquired servers fitted with the GB300 and other similar GB300 servers in Thailand. These were then “likely to [be used to] train its AI models,” he added. His comments stirred controversy in the industry as stakeholders eye Anthropic IPO with Moonshot also planning its IPO to rival global AI giants.
Kratsios Spotlights US Regulation of AI Infrastructure
The statement made a distinction between the practices used in the development of AI and the activity of Moonshot AI that the administration is accusing them of engaging in.
“The United States strongly supports the free and fair development of AI,” Kratsios said. The White House official noted that legitimate AI distillation plays a crucial role in innovation due to its ability to generate smaller and more efficient models.
However, he declared, “Large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable.”
The claims follow a recent easing of restrictions by the U.S. government on Anthropic’s most powerful AI systems, which occurred just weeks ago.
Anthropic said earlier this month that the U.S. Department of Commerce had lifted export restrictions on its most powerful models. After the ruling, the company stated it would reopen the international versions of Claude Fable 5 and Mythos 5 the following day.
Meanwhile, the White House has not provided any back-up proof to support Kratsios’ allegations. On the other hand, at the time of writing, Moonshot AI has not issued a public response to the allegations.
