American AI Companies Have Accused Chinese Counterparts Of Stealing. The Trump Admin Could Take Measures

Treasury Secretary Scott Bessent said the U.S. could sanction China over the "theft" of American AI models.

"If we see, especially that overseas models are stealing from our great companies, we have the ability to sanction them because of this theft," Bessent said in an interview with Fox News on Tuesday.

Bessent went o to say that "we are finding watermarks of our U.S. large language models on many of the Chinese models, and that's unacceptable." He added that the administration will "be looking at that in the coming days or weeks."

Anthropic recently accused Chinese technology giant Alibaba of conducting a large-scale campaign to extract capabilities from its Claude artificial intelligence models, alleging that operators linked to the company used thousands of fraudulent accounts to gain access to its systems.

In a June 10 letter sent to the U.S. Senate Committee on Banking, Housing, and Urban Affairs, Anthropic alleged that Alibaba carried out what it described as "the largest known distillation attack on Anthropic to date."

Anthropic said operators affiliated with Alibaba and its AI research division conducted approximately 28.8 million interactions with its models between April 22 and June 5 using roughly 25,000 fraudulent accounts. The company described the activity as an effort to extract information from its systems through a process known as distillation, a technique in which outputs from a more advanced model are used to train a smaller model.

"We believe combating the threat of illicit distillation requires coordinated action between government and industry," an Anthropic spokesperson said in a statement cited by CNBC. The company added that it would continue working with policymakers to protect what it described as American leadership in artificial intelligence.

The allegations also follow Anthropic's earlier claims involving other Chinese AI developers. In February, the company said it had identified what it described as industrial-scale distillation campaigns involving DeepSeek, Moonshot AI and MiniMax. Anthropic stated at the time that such efforts had become increasingly sophisticated and called for greater cooperation among technology companies, cloud providers and policymakers.

China, in turn, is considering reducing foreign access to its top AI models as well, according to another report.

Reuters detailed that authorities have held meetings with representatives of the country's main tech firms about the possibility, which would include models that have not been released yet. Leaks or the theft of proprietary technology would be considered an offense under the country's national security law, a source told the outlet.

Some of the companies that took part in the talks are Alibaba and ByteDance. Reuters noted that Chinese AI models have made strides in global adoption largely due to their lower costs and increasing capabilities. The curbing of access for companies overseas would drive up costs for many already relying on the technology.