![\"BizScoreAI,](\"https:\/\/bizscoreai.com\/wp-content\/uploads\/2026\/04\/BizScoreAI-Social-image.jpg\")
<\/a><\/p>\nBizScoreAI has the AI visibility scan scores how visible a business is to AI search and shows what its listing looks like to the engines people ask. Open the AI visibility scan<\/a>.<\/p>\n
The National Archives pulled an open-weight Alibaba Qwen AI search tool from the Federal Register website on Wednesday after social media users spotted the contradiction: a US government site was running a Chinese model while the FBI had just named Alibaba as a threat to American AI leadership. The model had been offered as an option for searching public comments on proposed regulations and was available for at least a day before being taken down.<\/p>\n
An archived snapshot of the Federal Register source code confirmed that the Qwen model was removed on Wednesday. A widely shared screenshot posted to X on September 15 by a user with the handle “tleilax___” showed the search option displayed on the government site. It is not clear when the National Archives, which runs the Federal Register, first offered the tool to visitors, and the agency has not commented on the removal.<\/p>\n
The White House and the FBI also did not comment.<\/p>\n
Earlier in September, the FBI named Alibaba among six leading Chinese firms it accused of conducting “industrial-scale distillation,” a practice the agency says helps China cut costs and shorten development time in the race for global AI leadership. That announcement made the appearance of an Alibaba product on a government website a glaring mismatch.<\/p>\n
Daniel Castro, president of the Information Technology and Innovation Foundation, called it an “insane” disconnect for a US agency to use an Alibaba model while the FBI encourages stakeholders to use only American models.<\/p>\n
The model in use was a small open-weight release identified as the Qwen3 0.6B level, according to a Chinese news report. It is not Alibaba’s largest flagship model. The report described it as serving solely to retrieve documents, providing no complex reasoning, and sending no government data to Alibaba or to any third party.<\/p>\n
For US agencies, downloading and running this kind of model locally can give the government more control over data than sending queries to larger models hosted externally. Security experts cited that ability, along with lower and more predictable pricing, as reasons such models have become a “default” AI search tool.<\/p>\n
Georgetown University Law Professor Anupam Chander noted that the Federal Register’s content is “already public, so the model was not working with sensitive government information.” Senator Mark Warner (D-Va.) said the security picture depends on whether any US data was processed by “Alibaba-controlled systems.”<\/p>\n
US Rep. John Moolenaar (R-Mich.), who chairs the House China Committee, said “no federal government entity should use a Chinese AI model.” “Doing so only makes the federal government more dependent on Chinese AI models, and that is not in the national interest,” Moolenaar said.<\/p>\n
Policy research submitted to the US-China Economic and Security Review Commission in March warned that the United States may already be at a disadvantage because it focused too heavily on maintaining a lead in frontier AI while failing to advance its own open-weight ecosystem. China, the researchers wrote, recognized that gap and moved to fill it.<\/p>\n
US export controls are calibrated to constrain frontier training by restricting access to advanced semiconductors. They do not address the small-model deployment cycle, which requires less advanced compute, draws on openly available base models, and generates advantage through application rather than pre-training.<\/p>\n
If the models that matter most for industrial AI are small, specialized, and open, the current US policy framework could be targeting the wrong layer of the competition. Companies like Nvidia and Meta have vowed to help the United States catch up. Researchers have warned that if the United States loses a broad user base to China, it could set back America’s ability to set the technical standards and norms that will govern AI development for years to come.<\/p>\n
Open-weight models are AI models whose trained parameters are published openly, so any organization can download them and run the model on its own hardware. Running the model locally means user queries never leave the operator’s own systems, which is one reason federal, state, and business users have gravitated to small open-weight releases for routine search and retrieval tasks.<\/p>\n
The Federal Register briefly used a small open-weight Alibaba Qwen model at the Qwen3 0.6B level to search public comments on proposed regulations. It is not Alibaba’s largest flagship model and does not provide complex reasoning.<\/p>\n
It was removed on Wednesday after social media users noticed the contradiction of a US government site using a Chinese AI model while the FBI had just named Alibaba among firms accused of industrial-scale distillation.<\/p>\n
Experts told Reuters it is unlikely the deployment posed any substantial risk because the site’s content is already public and the model was running locally rather than sending queries to Alibaba systems. Senator Mark Warner said the risk picture depends on whether any US data was processed by Alibaba-controlled systems.<\/p>\n
BizScoreAI has the AI visibility scan scores how visible a business is to AI search and shows what its listing looks like to the engines people ask. Open the AI visibility scan<\/a>.<\/p>\n<\/a><\/p>\n