Treasury Secretary Bessent Warns Chinese AI Firms of Sanctions Over Intellectual Property Theft Amid Open-Source AI Debate

U.S. Treasury Secretary Scott Bessent issued a stern warning on Wednesday regarding the use of American intellectual property by Chinese artificial intelligence companies. Speaking on the social media platform X, Bessent underscored that while the U.S. government supports open-source AI and the innovation it fosters, "open source is not open season on American IP."

Bessent specifically cautioned that Chinese firms engaging in "covert, industrial-scale distillation attacks" that cross the line into intellectual property theft could face sanctions and be placed on the Entity List, a U.S. trade blacklist that restricts access to American technology.

The Treasury Secretary’s comments come amid growing concerns over the unauthorized use of AI models and data. White House science chief Michael Kratsios also revealed on X that U.S. intelligence indicates Moonshot AI employed Anthropic’s Fable models to develop its Kimi K3 AI through large-scale distillation. Furthermore, Moonshot AI reportedly acquired or accessed NVIDIA GB300-equipped servers to train its AI models, raising alarms about potential IP violations.

In a related development, a company alleged that Chinese AI systems were improved through the exploitation of nearly 25,000 fraudulent accounts that generated more than 28.8 million interactions with the AI model Claude. This large-scale manipulation highlights the extent of covert efforts to leverage American AI technology illicitly.

The issue has drawn attention from U.S. lawmakers as well. Representative Ted Lieu (D-Calif.) publicly questioned the administration’s approach, criticizing the apparent contradiction between threatening sanctions over AI intellectual property theft and simultaneously permitting the sale of high-performance AI chips to China.

NVIDIA CEO Jensen Huang weighed in on the debate, stating that U.S. companies should “absolutely” use Chinese AI models. He described AI distillation as "fundamental to intelligence," suggesting a nuanced perspective on the technology’s cross-border development and use.

The Treasury’s warning signals a firm stance on protecting American intellectual property in the rapidly evolving AI sector. It also reflects broader geopolitical tensions surrounding technology transfer and national security.

Bessent’s remarks highlight the delicate balance between fostering open innovation and safeguarding proprietary technology. The U.S. government appears poised to take decisive action against entities that exploit open-source AI frameworks to conduct large-scale IP theft.

Meanwhile, market data shows that NVIDIA, a key player in AI hardware, maintains strong momentum and growth scores, underscoring its central role in AI development and the ongoing technology competition.

This evolving situation underscores the complex interplay between innovation, intellectual property rights, and international trade policies in the AI domain. The U.S. administration’s approach will likely influence future regulatory and enforcement measures concerning AI technology and cross-border data use.

As the AI landscape continues to expand, the Treasury’s message serves as a clear reminder that open-source does not equate to unrestricted access to American intellectual property, particularly when national security and economic interests are at stake.

The coming months may see increased scrutiny and potential sanctions targeting Chinese AI firms that engage in unauthorized use of U.S. technology, signaling a new phase in the global AI competition.