Key facts
- The Federal Register website briefly used Alibaba's Qwen AI search tool.
- The FBI accused Alibaba of "industrial-scale distillation" of US AI technology.
- The Qwen AI search tool was removed from the Federal Register website on Wednesday.
- The Federal Register website is run by the National Archives.
- The Qwen model used was a small, open-source version (Qwen3 0.6B level).
The US government removed an Alibaba AI search tool from the Federal Register website after its use was highlighted, raising concerns about potential contradictions with US law enforcement's stance on Chinese AI technology. The FBI had previously accused Alibaba and other Chinese firms of "industrial-scale distillation" of US AI models, a practice that could help American companies cut costs and development time in the global AI race.
The Federal Register website, operated by the National Archives, had briefly offered visitors the option to use Alibaba's Qwen AI model to search public comments on proposed regulations. The exact duration of its availability is unclear, but a screenshot shared on X on September 15 showed the option, and an archived version of the site's source code confirmed its removal on Wednesday.
Experts suggested that the deployment of the Qwen model, specifically a small, open-source version (Qwen3 0.6B level), likely posed no substantial security risks. This is because the content was already public, and the model was designed solely for document retrieval without processing sensitive government information or sending data to third parties. Some experts noted that such models, when run locally, could offer US agencies more control over data.
However, the situation highlighted a perceived disconnect, with some critics calling it "insane" for a US agency to use an Alibaba model while the FBI encourages reliance on American-made AI. US Representative John Moolenaar, chair of the House China Committee, stated that no federal entity should use Chinese AI models, citing national interest and dependence on Chinese technology.
The incident occurs amid a broader debate in the US about AI safety and regulation, and the potential competitive disadvantage the US might face if it restricts access to open-source models, including those from China. Policy research suggests that US export controls focused on advanced semiconductors may not address the growing importance of smaller, specialized AI models.