Federal Register Removed Alibaba-Powered Search Tool
The National Archives discontinued the tool following reports of alleged AI technology theft from Anthropic.
Updated on Sept. 30, 2026 in Artificial Intelligence

Live Poll
Should the U.S. government allow federal agencies to use artificial intelligence models developed in China?
The National Archives removed a search tool powered by Alibaba's open-weight Qwen AI model from the Federal Register website on September 24, 2026. This action followed public scrutiny and FBI accusations that the Chinese firm engaged in the malicious copying of AI technology from Anthropic.
Why it matters
The incident highlights growing national security concerns regarding the deployment of foreign-developed AI models on U.S. federal infrastructure. It underscores the difficulty of managing the security risks associated with open-weight models that can be easily integrated into domestic systems.
The Federal Register previously utilized an integration of Qwen, an open-weight AI model designed to be deployed on custom infrastructure. Unlike closed-source APIs, open-weight models are downloaded and modified by developers to power localized applications.
The players
National Archives
The agency responsible for managing federal records and operating the Federal Register website.
Alibaba
A major Chinese technology conglomerate whose Qwen AI models are at the center of intellectual property disputes.
Anthropic
A prominent AI research company that develops large language models and accused Alibaba of unauthorized technology copying.
FBI
The U.S. domestic intelligence and security service currently investigating the claims of intellectual property theft.
The details
Open-weight models—AI systems where the internal parameters are made publicly available for download—allow organizations to host and run the software on their own private servers. This flexibility facilitates rapid integration, though it complicates oversight as entities can deploy and customize the underlying code without needing direct permission from or constant connectivity to the original developer.
Timeline
April 2026: House committees issued an inquiry letter to Airbnb regarding Qwen.
September 23, 2026: The search tool was still available on the Federal Register website.
September 24, 2026: The search tool was removed from the Federal Register.
The Tech Race
This development follows a pattern of heightened federal scrutiny regarding the standards established by Executive Order 14110 on Safe, Secure, and Trustworthy AI. It reflects a broader shift toward tighter vetting of foreign-origin AI models used in critical government infrastructure.
Users of the Federal Register website will no longer have access to the specific Qwen-based AI search features for retrieving document information. The impact is limited to this specific site interface, as the underlying model remains available to developers and other third-party platforms.
The takeaway
The incident demonstrates the sensitivity of government digital infrastructure to foreign AI sourcing. Future developments to watch include potential federal guidance on the approval process for third-party AI integrations within official archives.
Further reading
For more on the national security implications of AI deployment, visit the Artificial Intelligence section.
Source note: This article includes information reported by The Times of India.
Live Poll
Should the U.S. government allow federal agencies to use artificial intelligence models developed in China?





