WASHINGTON, July 22 (Reuters) – The United States has accused China's Moonshot AI of stealing technology from Anthropic's advanced large language model, Fable, to develop its latest K3 model. Michael Kratsios, a top Trump administration technology official, said Moonshot engaged in large-scale industrial distillation to extract proprietary U.S. technology. He added that the company had also obtained Nvidia GB300 chips and used them in Thailand to train its models. U.S. Treasury Secretary Scott Bessent said he was considering adding Moonshot AI to a trade blacklist and imposing sanctions.

The accusations come amid escalating tensions between Washington and Beijing over artificial intelligence and technology security. A Chinese embassy spokesperson in Washington dismissed the claims as unfounded, asserting that China respects intellectual property rights. Anthropic’s public policy executive Sarah Heck warned that such alleged theft poses serious national security risks for the United States.

The dispute follows earlier reports that several Chinese AI firms, including Moonshot, used Anthropic’s Claude model through fake accounts to enhance their own systems. Moonshot recently launched its Kimi K3 model, claiming it is the world’s largest open-weight AI system, rivaling Anthropic’s Fable.