Federal authorities in the United States have formally charged half a dozen Chinese artificial intelligence developers with orchestrating large-scale theft of proprietary technology from American AI companies through a technique known as 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
This coordinated announcement was released by the National Security Agency, Federal Bureau of Investigation, and Cybersecurity and Infrastructure Security Agency on Tuesday, September 8.
The distillation process involves training a compact AI system by utilizing responses generated from a more sophisticated, resource-intensive model. When conducted with proper authorization, this represents an accepted methodology within the technology sector.
Federal intelligence officials assert that Chinese enterprises employed this approach without obtaining consent, leveraging various access points to American AI infrastructure while circumventing usage agreements.
According to the official advisory, these operations were executed “at an industrial scale” and presumably occurred with Beijing’s awareness.
The half-dozen enterprises identified are DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI. China’s diplomatic mission has remained silent on comment requests. Spokespeople for the accused corporations similarly have not issued responses.
American technology companies allegedly affected encompass Anthropic, OpenAI, Google, and SpaceX. Intelligence agencies stated Chinese developers leveraged the appropriated information to enhance capabilities across domains including mathematical computation and software code analysis.
Anthropic had formally complained about Alibaba in June, alleging the company deployed thousands of fraudulent user accounts to gain unauthorized entry to its Claude AI platform.
OpenAI has similarly voiced objections regarding Chinese rival firms obtaining improper access to its technology infrastructure.
Federal authorities have recommended American AI companies implement urgent protective measures, including modifying system responses when distillation attempts are detected and establishing information-sharing protocols among industry participants.
The Trump administration had previously committed in April to confronting this challenge. Congressional representatives are simultaneously examining legislative options to impose penalties on Chinese enterprises discovered replicating American AI technology.
Treasury Secretary Scott Bessent indicated in July that economic sanctions might be deployed against any Chinese company determined to be misappropriating intellectual property.
This advisory arrives several weeks ahead of a scheduled diplomatic summit between President Trump and President Xi Jinping in Washington during late September.
Beijing’s foreign affairs department has dismissed these allegations. Ministry spokesperson Mao Ning characterized China’s AI advancements as products of domestic scientific innovation and urged Washington to cease what she termed unfounded accusations.
Both nations are additionally organizing a dedicated AI safety discussion scheduled for mid-September.
Reuters previously disclosed in July that Chinese military research teams utilized outputs from American AI systems to enhance China’s defense technology programs.
This most recent statement from US security apparatus may accelerate congressional efforts to enact legislation addressing these practices.
The post Six Chinese AI Companies Face Technology Theft Accusations From US Intelligence Agencies appeared first on Blockonomi.
