TechNewsReel
Live

US Agencies Accuse Chinese AI Firms of Industrial-Scale Model Theft

The FBI, NSA, and CISA allege that companies including DeepSeek and Alibaba distilled billions of tokens from American models to build their own systems.

TechNewsReel Newsroom · September 9, 2026

Three U.S. federal agencies have accused several Chinese artificial intelligence companies of conducting "industrial-scale" theft of American AI models. In a joint advisory issued on September 8, 2026, the FBI, NSA, and CISA claimed that Chinese firms have systematically stripped capabilities from U.S. frontier models to accelerate their own development.

According to the advisory, companies including DeepSeek, Alibaba, and Moonshot AI—along with MiniMax, StepFun, and Z.AI—have distilled billions of tokens from U.S. models since at least late 2024. The U.S. government alleges this process was not merely a supplement to original research but served as a primary driver for the development of these systems. U.S. Treasury Secretary Scott Bessent dismissed the threat to American leadership, stating, "The Chinese distill our models and they can never get ahead of us."

The Distillation Dispute

The conflict centers on "distillation," a technical process where a smaller, more efficient AI model is trained using the outputs of a larger, more advanced "teacher" model. While the U.S. government characterizes the scale and coordination of these efforts as intellectual property theft, Chinese officials argue the practice is a standard industry technique used globally.

China's Ministry of Foreign Affairs and its Commerce Ministry have dismissed the U.S. allegations as "slander" and "unfounded accusations." Spokesperson Mao Ning asserted that China’s progress in the field is the result of "high-level technological self-reliance and strength" rather than theft.

Geopolitical Consequences

This escalation marks a significant hardening of the technological race for AI dominance between the two superpowers. The dispute highlights growing friction over AI sovereignty as the U.S. seeks to protect its intellectual property and maintain a competitive edge in frontier model development.

For the industry, the accusations could lead to more aggressive crackdowns from the U.S. government on Chinese AI firms, including further sanctions or restrictions on the flow of data and hardware. Such tensions threaten to fragment the global AI ecosystem into two distinct, competing blocs with little interoperability.

Potential Retaliation

Beijing has signaled that it will not remain passive in the face of U.S. pressure. China's Commerce Ministry warned that if the U.S. suppresses Chinese AI companies under the pretext of targeting distillation, China will take "resolute countermeasures."

While the U.S. continues to monitor the extent of the distillation efforts, it remains to be seen whether these accusations will lead to formal legal actions or new trade barriers. For now, the two nations remain locked in a cycle of accusation and denial, with the definition of "industry practice" versus "theft" remaining a central point of contention.

Sources

Get a notification when a big story breaks. A few a day at most — no spam.