US threatens sanctions on China’s AI models for stealing IP

On Tuesday, Treasury Secretary Scott Bessent said the US will examine open source models from China for signs of intellectual property theft, threatening sanctions against Chinese AI companies if IP theft is established.

“We’ve seen a lot of talk about open source models coming and threatening major language brands in the US,” Bessent said on Fox Business on Tuesday. “This administration supports open source models, but we do not support IP theft. If we see, in particular, that overseas models are stealing from our big companies, we have the ability to punish ourselves for this crime.”

Bessent’s comments were first reported by Bloomberg.

This statement comes as Chinese models – most recently Moonshot AI’s Kimi K3 – are growing in skill and popularity, threatening to undermine the business models of top American AI companies such as OpenAI and Anthropic, and their ability to raise additional capital to continue developing frontier models.

On Monday, Axios reported that the Trump administration is considering a total ban on China’s open source models, though others have disputed that claim.

AI companies have been warning for months about campaigns by foreign actors to copy their AI technology and reuse it as open source. In April, the White House said it would work more closely with AI companies to fight crime.

Sanctions from the US against Chinese models will add to the growing list of strategies the government is trying to maintain its lead in the AI ​​race. After limiting China’s access to advanced chips and tightening export controls, Washington is now signaling it may target AI models itself, a move that could mark a major escalation in technological competition between border labs and China’s open source alternatives.

The distillation model is a technique that allows some of the capabilities of a larger model to be translated into a smaller system that works more easily – but not everyone agrees that copying another company’s model is stealing.

Earlier this month, Microsoft CEO Satya Nadella criticized the big labs for making this assumption: “While a positive change from model providers who have the right to use the right to train models on public data is necessary, I find it strange that the current situation is to turn around and put restrictive policies on distillation.”

The training processes of AI labs continue to be a source of legal risk for companies. Anthropic this week got the green light to start cutting authors’ checks as part of its $1.5 billion settlement after a judge ruled it illegally downloaded and retained millions of copyrighted books to train its AI.

In addition, some in the industry argue that distillation is not the only reason for China to acquire US AI companies.

“We know that watering is a very small factor in the ability to create good models, and it’s a practice that everyone does, including US companies,” said Hugging Face CEO Clem Delangue on a recent episode of TechCrunch’s Equity podcast. “If it were easy to do a distillation to find the good in building AI models, there would be many other countries, including the US, that have the best open source AI. The fact is that they have really good research groups in China…that take a much more open and collaborative approach to AI than the US”