DHS bans DeepSeek from federal civilian agencies
Original: DeepSeek banned from U.S. federal civilian agencies under new DHS order
Why This Matters
Federal AI procurement rules are taking shape; this sets a precedent for Chinese model bans.
The U.S. Department of Homeland Security has ordered federal civilian agencies to ban DeepSeek AI from government IT systems, citing national security concerns over the Chinese-developed model.
The Department of Homeland Security has issued a directive prohibiting DeepSeek's AI models from use across U.S. federal civilian agency IT systems. The order follows mounting concern in Washington over Chinese-developed AI tools and their potential to expose sensitive government data or provide Beijing with intelligence advantages. DeepSeek, developed by a Chinese hedge fund-backed lab, drew global attention in early 2025 after its R1 model matched or exceeded leading U.S. models at a fraction of the reported training cost. That debut sparked both admiration and alarm in Washington. The DHS ban extends a pattern of federal action against Chinese tech: TikTok and Huawei faced similar restrictions in prior years. No detailed technical justification was published alongside the order, per Reuters' reporting, but the directive applies to civilian agencies under DHS oversight. Congress had already seen bipartisan calls to restrict DeepSeek on government devices before any formal order was issued.