US threatens sanctions on Chinese AI models over IP theft

Treasury Secretary Scott Bessent said Tuesday that the U.S. will examine open source AI models from China for signs of intellectual property theft, and threatened sanctions against Chinese AI companies if such theft is established. “This administration supports open source models, but what we do not support is IP theft,” Bessent said on Fox Business. “If we see, especially, that overseas models are stealing from our great companies, we have the ability to sanction them because of this theft.” His comments were first reported by Bloomberg.

The statement comes as Chinese models — most recently Moonshot AI’s Kimi K3 — gain in capabilities and popularity, threatening to erode the business models of top American AI firms like OpenAI and Anthropic and their ability to raise more capital for frontier-model development. Separately, Axios reported Monday that the Trump administration is considering a wholesale ban on Chinese open source models, though others have disputed that claim.

AI companies have warned for months that foreign actors are copying their technology and redeploying it as open source; in April, the White House said it would work closely with AI firms to combat the theft. Sanctions would add to a growing list of strategies the U.S. government is using to defend its AI lead, following restrictions on China’s access to advanced chips and tighter export controls — now potentially extending to the models themselves. Much of the dispute centers on model distillation, a technique that transfers some of a larger model’s capabilities into a smaller, cheaper-to-run system; not everyone agrees that distilling another company’s model amounts to theft, and Microsoft CEO Satya Nadella has criticized large labs for treating it as such while relying on broad fair-use rights to train their own models on public data.

Separately this week, Anthropic got the green light to start cutting checks to authors as part of its $1.5 billion settlement, after a judge ruled the company had illegally downloaded and stored millions of copyrighted books to train its AI. Hugging Face CEO Clem Delangue has argued that distillation is only a minor factor behind China’s progress, pointing instead to strong Chinese research teams taking a more open, collaborative approach to AI than their U.S. counterparts.

Why it matters

If enacted, sanctions on AI models themselves — beyond chips and export controls — would mark a significant escalation in the U.S.-China technology competition, raising the stakes for any American firm that uses or builds on open source models with Chinese origins.