US Treasury Secretary Scott Bessent warned that the government could impose sanctions on Chinese artificial intelligence companies if evidence emerges that they have stolen intellectual property from American AI developers.

Speaking on Fox Business, Bessent emphasized that while the Trump administration supports open source AI, it will act against foreign companies using stolen US technology. "If we see, especially, that overseas models are stealing from our great companies, we have the ability to sanction them because of this theft," he stated.

Context and Industry Reactions

The caution comes amid rising competition from Chinese AI models like Moonshot AI’s Kimi K3, which are gaining traction in coding and autonomous tasks where firms like OpenAI and Anthropic have dominated. Axios reported that the Trump administration has been considering broader restrictions on Chinese open source AI models, though this report has been disputed.

Sanctions would escalate Washington's efforts to secure an AI advantage, building on previous measures that restricted China’s access to advanced semiconductors and tightened export controls. Targeting software developers directly would extend US restrictions beyond hardware to include training methodologies.

The core dispute revolves around model distillation, a process transferring capabilities from larger to smaller AI systems. US companies accuse foreign developers of accelerating their own AI by exploiting these methods improperly, while critics argue this is a standard practice in AI development rather than theft.

Microsoft CEO Satya Nadella recently challenged the claim that distillation constitutes intellectual property theft, pointing out that major AI firms rely on fair use doctrines when training models on public data but impose strict conditions on others attempting similar techniques. Similarly, Hugging Face CEO Clem Delangue attributed China's AI progress more to solid research teams and open development approaches than to distillation.

The debate occurs as US firms face their own legal challenges regarding training data. Anthropic received court approval to start disbursing $1.5 billion in settlements with authors after a judge found the company had improperly downloaded copyrighted material.

This article provides information only and is not financial advice.