Saturday
Sep 19, 2026
Socialloop
Photo by Albert Stoynov on Unsplash
AI Policy & Ethics

US government website used Chinese model the FBI called "malicious"

Ars TechnicaSeptember 18, 202690% confidence

The United States National Archives recently deactivated an artificial intelligence search function on the Federal Register website after observers pointed out its reliance on an open-source model developed by the Chinese technology giant Alibaba.

This discovery triggered immediate criticism from cybersecurity experts and political figures because federal law enforcement agencies had previously cautioned domestic organizations against using Chinese-developed AI systems, accusing foreign firms of copying Western innovations to accelerate their own capabilities. Despite the swift removal of the search tool, representatives from the National Archives, the White House, and the Federal Bureau of Investigation have declined to publicly comment on how the system was initially deployed or why it was subsequently withdrawn.\n\nAlthough some lawmakers expressed concern over potential security vulnerabilities and an increased reliance on foreign technology, industry specialists suggested the actual risk in this specific instance was negligible. The feature utilized a highly compact, open-source version of Alibaba’s Qwen model that was hosted locally on government infrastructure to retrieve public records, meaning no sensitive or classified governmental data was transmitted externally. Nevertheless, the incident has intensified a broader domestic debate regarding international technology competition and regulatory strategy. Policymakers are increasingly questioning whether American export controls focus too heavily on restricting cutting-edge frontier systems while inadvertently leaving the United States lagging in the rapidly expanding sector of smaller, highly accessible open-source models.

Summary generated September 18, 2026. AI summaries can make mistakes.

Read Original on Ars Technica

Category

Topic (AI-estimated)

AI & Machine Learning

90% confidence


AI Policy & Ethics

This category is an AI-estimated classification based on the article's content and may not be fully accurate.

Sentiment

Sentiment

Neutral

Recent Posts

Popular Tags

AI & Machine LearningAI Policy & Ethics

We use cookies for essential site functionality and, with your consent, to understand how the site is used.