US accuses six Chinese AI firms of IP theft via distillation

US escalates AI intellectual property dispute with China

The United States government has formally accused six Chinese artificial intelligence companies of systematically copying proprietary technology from American AI firms through a process known as distillation, according to a statement released Tuesday by US law enforcement and intelligence officials.

The named companies include DeepSeek, Moonshot AI, and Alibaba, which owns the South China Morning Post. Officials allege these firms used variants of AI models developed by Anthropic, OpenAI, Google, and SpaceX to train their own products more rapidly and at lower cost.

Distillation involves training smaller, more efficient AI models by using the outputs of larger, more expensive systems as training data. While the technique itself is not inherently illegal, US officials characterized the Chinese companies' activities as "aggressive, malicious and targeted" and conducted "at an industrial scale."

Timing and geopolitical context

The accusations arrive at a diplomatically sensitive moment. Chinese President Xi Jinping is scheduled to visit the United States later this month for meetings with President Donald Trump. Additionally, the two nations are preparing for a mid-September dialogue focused on AI safety risks.

This marks the second time this year the US has leveled similar allegations against Chinese AI developers. Officials made comparable claims in April, prior to Trump's visit to Beijing.

US officials stated the distillation activities occurred "likely with Chinese government awareness," though they stopped short of alleging direct state direction. The Chinese embassy in Washington had not responded to requests for comment at the time of the report.

Why it matters

The distillation allegations represent a new front in US-China technology competition, moving beyond export controls and chip restrictions to focus on how AI models themselves can be replicated. US officials argue the practice not only reduces research costs for Chinese competitors but could enhance China's "military and cyberattack capabilities that could be used against the US and our allies." The timing suggests the administration may be establishing negotiating leverage ahead of high-level diplomatic engagements, though the accusations could equally complicate those discussions. For AI companies, the dispute highlights growing scrutiny over model access and the potential weaponization of openly available AI outputs.

Military implications cited

The Tuesday statement went beyond commercial concerns, asserting that the alleged intellectual property theft could strengthen China's military and cyber capabilities. This echoes previous reporting that Chinese military researchers have utilized outputs from leading American AI models to advance domestic defense systems.

These details were first reported by Reuters and the South China Morning Post.