US Homeland Security bans DeepSeek from federal civilian agencies’ systems and purchases
The Department of Homeland Security has ordered federal civilian agencies to stop buying products from China’s DeepSeek and to remove DeepSeek technology, including third-party services built on its models, from their IT systems.
This story was prepared automatically by software.
The US Department of Homeland Security (DHS) has ordered federal civilian agencies to exclude Chinese AI company DeepSeek from their purchases and to remove its technology from their systems. The order was made public on October 19 and is scheduled to be published in the Federal Register on October 20.
What the order covers
The order has two parts. Civilian agencies must exclude information technology products and services developed by Hangzhou DeepSeek Artificial Intelligence Co. and its affiliates from future procurements, and they must remove all DeepSeek and DeepSeek-affiliate products, as well as third-party products and services built using DeepSeek implementations, from their IT systems.
The order was issued under the Federal Acquisition Supply Chain Security Act of 2018 (FASCSA), following a review of a recommendation by the Federal Acquisition Security Council. It applies to civilian agencies such as the Treasury, State Department and Justice Department. Defense Department, national security and intelligence community systems fall outside it, as those are covered by the Secretary of Defense and the Director of National Intelligence.
The risks DHS cites
According to CyberScoop, Homeland Security Secretary Markwayne Mullin wrote that DeepSeek’s models present “multiple threat vectors” to federal civilian IT systems. He said DeepSeek stores data on servers in China, where national security laws give the government wide latitude to obtain data from domestic companies; that models hosted by DeepSeek collect personally identifiable information and trace users’ “digital fingerprints”; and that the prompt-and-response process “may result in the transfer of sensitive data” to the Chinese government. Combined with publicly available data, that information “could produce actionable intelligence on U.S. persons and/or government programs,” Mullin wrote.
The documents indicate the ban has effectively been in place for months: CyberScoop notes that Mullin’s order letter is dated June 26. Agencies may request exceptions, but must do so 60 days in advance with a “compelling justification.” Until DeepSeek is fully removed, agencies are encouraged to monitor traffic in and out of their networks, tighten logging and data-exfiltration controls, and flag anomalous activity.
Another front in the US-China AI rivalry
As the BBC noted, the move follows criticism of Chinese AI from the Trump administration and from American firms. Anthropic and OpenAI have both accused Chinese companies, including DeepSeek, of “distilling” their models to copy capabilities. DeepSeek’s low-cost open models remain popular with startups and are incorporated into products from US companies such as Snowflake, Vercel and OpenRouter; the order applies only to federal civilian agencies.
The Wall Street Journal also reported the order. The BBC said DHS, DeepSeek and the Chinese embassy in Washington had been contacted for comment.
Sources: BBC, CyberScoop, The Wall Street Journal, Federal Register. Image: CFOTO/Future Publishing via Getty Images (via BBC).
Related articles
