Bessent says U.S. could sanction China over AI model 'theft'

U.S. Treasury Secretary Scott Bessent announced the Trump administration will investigate whether Chinese AI models have been distilled from American models, threatening sanctions if evidence of intellectual property theft is found. The statement comes as Chinese open-weight models, particularly Moonshot AI's Kimi K3, outperform leading U.S. offerings from OpenAI and Anthropic on industry benchmarks. Bessent cited watermarks of U.S. large language models on Chinese models as unacceptable. Anthropic previously alleged Alibaba conducted the largest known distillation attack. The U.S. and China plan to hold AI talks in September 2026, with Bessent representing the U.S. Both Anthropic and OpenAI have faced their own accusations of theft, including copyright lawsuits from authors and The New York Times.