U.S. Treasury Secretary Scott Bessent announced that the Trump administration will investigate whether Chinese artificial intelligence (AI) models have been distilled from American models, raising the possibility of sanctions if intellectual property (IP) theft is confirmed [1]. Bessent specifically referenced the growing competitiveness of Chinese open-weight models, such as Moonshot AI's Kimi K3, which was released earlier this month and reportedly outperforms leading American offerings from companies like OpenAI and Anthropic on some industry benchmarks [1].
Bessent explained that 'distillation' is a technical method where a smaller AI model is trained using outputs from a more capable model, and he warned, 'If we see, especially that overseas models are stealing from our great companies, we have the ability to sanction them because of this theft' [1]. Concerns about Chinese AI practices have been heightened by a letter from Anthropic to the U.S. Senate Committee on Banking, Housing, and Urban Affairs, alleging that Alibaba carried out 'the largest known distillation attack' against Anthropic to date [1].
The U.S. and China are scheduled to hold talks on AI in September, with Bessent representing the U.S. side, according to a Reuters report cited by CNBC [1]. Bessent stated, 'We are finding watermarks of our U.S. large language models on many of the Chinese models, and that's unacceptable,' emphasizing that the administration will be investigating these issues in the coming days or weeks [1].
The article also notes that both Anthropic and OpenAI have faced their own allegations of IP theft. Anthropic agreed to pay $1.5 billion in September to settle a class action lawsuit with authors who claimed the company illegally downloaded books from pirated databases, while The New York Times sued OpenAI and Microsoft in 2023 for alleged copyright infringement in model training [1].
CONCLUSION
The U.S. government is intensifying scrutiny of Chinese AI development practices, with potential sanctions on the table if IP theft is confirmed. This development signals heightened tensions in the global AI race and could have significant implications for cross-border technology competition and regulation.