US threatens sanctions against Chinese AI models over IP theft - TechCrunch

What the report says

TechCrunch reported that U.S. Treasury Secretary Scott Bessent said on Tuesday that Washington may scrutinize Chinese open-source AI models for possible intellectual property theft and could impose sanctions on Chinese AI companies if theft is found. Speaking on Fox Business, Bessent said the Trump administration backs open-source models but would not tolerate overseas systems taking technology from U.S. firms.

The comments, first reported by Bloomberg and covered by TechCrunch, come as Chinese AI systems gain attention for improving performance and wider adoption. TechCrunch cited Moonshot AI’s Kimi K3 as a recent example and noted that such models could pressure leading U.S. AI companies including OpenAI and Anthropic, whose business models and fundraising depend partly on maintaining a technological edge.

The report places Bessent’s warning within a broader U.S. strategy to slow China’s AI progress, following restrictions on advanced chips and tighter export controls. TechCrunch also noted that Axios reported the administration was considering a broader ban on Chinese open-source models, though that claim has been disputed. The White House said in April it would work with AI companies to address technology theft concerns.

A key unresolved issue is how to define theft in AI development. TechCrunch highlighted debate over model distillation, a method for transferring some capabilities from a larger model into a smaller one. Some companies view unauthorized distillation as improper copying, while others argue the practice is common and not the main reason Chinese models are improving. The debate matters because sanctions targeting models themselves would be a notable escalation in U.S.-China technology competition.