US Government Accuses Chinese AI Firms of Industrial Scale Intellectual Property Theft

US federal agencies and officials accused Chinese artificial intelligence companies on Tuesday of maliciously copying technology from American AI firms. The FBI, National Security Agency, and Cybersecurity and Infrastructure Security Agency alleged that firms including Alibaba and DeepSeek used a technique called distillation to extract proprietary functionalities from US frontier AI models. Distillation involves training smaller AI models using the output of larger, more expensive ones. While the technique can be used for legitimate research, US officials claim these campaigns are conducted at an industrial scale to reduce research costs and accelerate the time to market.

Treasury Secretary Scott Bessent stated that sanctions and Entity List designations are possible for firms that cross the line into intellectual property theft. He compared the practice to copying homework and argued that China can never get ahead of the US in AI because of this reliance on American models. Other US officials noted that these activities likely occur with the awareness of the Chinese government and could enhance China's military and cyberattack capabilities.

Anthropic specifically alleged that three Chinese labs used approximately 24,000 fraudulent accounts to generate over 16 million exchanges with its model, Claude. To bypass geographic restrictions and safeguards, the US agencies claimed Chinese firms use a gray market of proxies known as transfer stations. These allegations come ahead of a planned visit by Chinese President Xi Jinping to the US in late September, where AI safety and technology are expected to be discussed.

Framing of the event varied by political lean. Center left sources emphasized the industrial scale of the theft and the potential risks to national security. Center right sources focused on the comments of Treasury Secretary Bessent regarding China's inability to surpass US technical leadership.