Key facts
- The U.S. is considering sanctions against Chinese AI firms for alleged intellectual property theft.
- Treasury Secretary Scott Bessent stated that the U.S. supports open-source models but not IP theft.
- Concerns exist that Chinese developers are using stolen U.S. technology and model weights to create competitive AI systems.
- U.S. officials have reportedly found 'watermarks' of U.S.-made AI models within Chinese models.
- Potential sanctions could include asset freezes and restrictions on financial system access for targeted entities.
- The U.S. is also examining adding Chinese AI labs to the Commerce Department's 'entity list'.
The U.S. is preparing to impose potential sanctions on Chinese artificial intelligence developers over allegations of intellectual property theft, according to Treasury Secretary Scott Bessent. Bessent stated on Fox Business that while the administration supports open-source AI models, it will not tolerate IP theft, particularly from overseas entities using stolen U.S. technology.
The warning comes as Chinese open-weight AI models have gained significant traction, rivaling the capabilities of leading U.S. platforms like OpenAI's GPT-4 and Anthropic's Claude. U.S. officials are questioning how China has rapidly closed the development gap, with some suggesting it stems from unauthorized use of U.S. proprietary data, model weights, or architectural innovations.
Bessent indicated that U.S. officials have found evidence, described as 'watermarks,' of U.S.-made AI models within Chinese systems, deeming this practice "unacceptable." The potential sanctions could escalate enforcement beyond current export controls, possibly involving asset freezes and exclusion from the U.S. financial system for targeted Chinese AI companies and researchers. The U.S. is also reportedly examining options such as adding Chinese AI labs to the Commerce Department's 'entity list.'
Concerns have been amplified by the performance of Chinese models like Moonshot's Kimi K3 and Zhipu AI's GLM-5.2, which offer advanced capabilities at a lower cost. This development challenges the previous assumption of a significant U.S. lead in large language model development. High-level U.S.-China talks on artificial intelligence are slated for September, where these issues are expected to be a central focus.