TLDR
- The NSA, FBI, and CISA jointly accused six Chinese AI companies of using “distillation” to steal from US AI models at an industrial scale.
- DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI were named in the statement.
- The Chinese firms allegedly targeted models from Anthropic, OpenAI, Google, and SpaceX.
- US agencies say the activity likely happened with Chinese government knowledge.
- The accusations come weeks before a planned summit between Trump and Xi Jinping.
The US government has accused six Chinese artificial intelligence companies of systematically copying technology from American AI firms using a method called distillation.
🚨BREAKING: The NSA, FBI and CISA have jointly accused DeepSeek, Alibaba and four other Chinese AI firms of stealing from US models "at an industrial scale."
The agencies say the companies used distillation to extract proprietary knowledge from American AI systems since 2024,… pic.twitter.com/hHRAxweptv
— Coin Bureau (@coinbureau) September 9, 2026
The joint statement came from the National Security Agency, the FBI, and the Cybersecurity and Infrastructure Security Agency on Tuesday, September 8.
What Is Distillation and Why Does It Matter
Distillation is a technique where a smaller AI model is trained using outputs from a larger, more powerful one. When done with permission, it is a standard industry practice.
The US agencies say Chinese companies used this method without approval, accessing American AI systems through multiple pathways to get around terms of service.
The statement said this activity happened “at an industrial scale” and likely with the knowledge of the Chinese government.
The six companies named are DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI. The Chinese embassy did not respond to requests for comment. Representatives for the named companies also did not respond.
The targeted American companies include Anthropic, OpenAI, Google, and SpaceX. The agencies said Chinese firms used the stolen knowledge to improve abilities in areas like math problem-solving and code review.
Anthropic and OpenAI Had Already Raised Concerns
Anthropic had previously accused Alibaba in June of using thousands of fake accounts to access its Claude model without authorization.
OpenAI has also raised similar complaints about Chinese competitors gaining access to its systems.
US agencies advised American AI developers to take immediate steps, including changing how their systems respond to suspected distillation attempts and sharing intelligence with each other.
The Trump administration had already vowed in April to address the issue. US lawmakers are also looking at measures to penalize Chinese companies found copying from American AI firms.
US Treasury Secretary Scott Bessent warned in July that sanctions could follow for any Chinese firm found to be stealing intellectual property.
The statement was issued weeks before President Trump is set to meet Chinese President Xi Jinping in Washington in late September.
China’s foreign ministry rejected the claims. Spokesperson Mao Ning said China’s AI progress is the result of its own scientific work and called on the US to stop what she described as false accusations.
The US and China are also preparing for a separate AI safety dialogue planned for mid-September.
Reuters had previously reported in July that Chinese military researchers used outputs from US AI models to advance China’s defense capabilities.
The latest statement from US security agencies may push lawmakers to move faster on legislation targeting the practice.
Stop guessing and start investing with confidence. KnockoutStocks gives you the AI insights, market intelligence, and stock research you need to spot opportunities, cut through the noise, and make smarter investment decisions — all in one powerful platform.
Sign up today and get 50% OFF full access to our premium stock picks.
Simply use coupon code SPECIAL50 at checkout to claim your exclusive discount.







