The US Department of Homeland Security (DHS) has banned the use of DeepSeek across federal agencies. This directive applies to new procurements, existing systems, and third-party services that integrate DeepSeek technology. Officials cited concerns that Chinese law could grant the government in Beijing access to sensitive data belonging to US federal employees.
These new measures expand upon restrictions established in December 2025, which initially targeted the use of DeepSeek within national security and intelligence systems. While the ban now covers federal civilian agencies, government entities may apply for specific exemptions to these requirements.
The scrutiny intensified following the January 2025 release of the R1 reasoning model, which gained traction due to its low cost and open-weight architecture. In September, the US accused DeepSeek and five other Chinese firms, including Alibaba and Moonshot AI, of using model distillation to extract data from American systems.
The Chinese Ministry of Foreign Affairs, represented by Mao Ning, criticized the US decision, labeling it a politicization of trade and technology. Meanwhile, reports indicate that China is considering export restrictions on 23 AI-related technologies, including advanced chip designs and specific AI models, as the technological rivalry between the two nations continues.
