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

• Bessent said the U.S. has the ability to sanction foreign AI companies found stealing from American models through distillation, a training method where smaller models are built using outputs from stronger existing models
• Chinese startup Moonshot AI recently released a model that outperforms OpenAI and Anthropic across some industry benchmarks, raising concerns about the durability of U.S. leadership in AI
• The U.S. and China are planning to hold trade talks in September, with Bessent representing the U.S., and the administration expects to address the AI model theft issue within days or weeks

Market Summary: U.S. Considers Sanctions on China Over AI Model Theft

Key Developments

U.S. Treasury Secretary Scott Bessent announced Tuesday that the Trump administration will investigate whether Chinese AI models have stolen intellectual property from American companies through a process called "distillation." The government is prepared to impose sanctions if theft is confirmed.

Technical Details

Distillation involves training smaller models using outputs from more advanced existing models. OpenAI submitted testimony to the Senate Banking Committee last month alleging that Chinese tech company DeepSeek conducted "the largest known distillation attack" against it. Bessent stated that U.S. watermarks have been discovered in multiple Chinese large language models.

Competitive Landscape

Chinese open-weight AI models are increasingly competitive with leading American offerings from OpenAI and Anthropic. Chinese startup Moonshot AI released its new model earlier this month, which outperforms some U.S. models on industry benchmarks, raising concerns among business leaders and government officials about America's AI dominance.

Market Implications

The potential sanctions represent an escalation in U.S.-China tech tensions and could impact global AI development and competition. According to Reuters, the U.S. and China are scheduled to hold discussions in September, with Bessent representing the U.S. side.

Timeline

Bessent indicated the investigation will proceed "in the coming days or weeks," suggesting imminent action on this issue.

This development affects major AI companies including OpenAI, Anthropic, and Chinese firms like DeepSeek and Moonshot AI, with broader implications for the technology sector and U.S.-China trade relations.