US Orders Federal Agencies to Purge DeepSeek AI Tools
US orders civilian agencies to remove DeepSeek and third-party tools using its AI, citing data security risks and potential access by Chinese authorities.
Summary
As of October 20, 2026, the US has ordered federal civilian agencies to stop buying DeepSeek products and services and remove them from existing IT systems, citing national security concerns tied to the Chinese AI startup’s operations in China. The directive covers third-party products and services built on DeepSeek technology, although agencies may request exceptions. DHS warned that Chinese law could give Chinese authorities access to sensitive US government user data. It expands a December 2025 law restricting DeepSeek on intelligence national security systems to civilian agencies, new purchases and installed technology.
DeepSeek’s low-cost, open-weight models compete with OpenAI and Meta. Its R1 reasoning model drew global government and regulatory scrutiny after triggering a sharp Nvidia share sell-off in January 2025. A September report accused DeepSeek and five other Chinese AI companies, including Alibaba and Moonshot AI, of using large-scale model distillation to extract information from American AI systems. Earlier in 2026, OpenAI told US lawmakers it had observed suspected DeepSeek attempts to distill its models, while Anthropic identified unauthorized efforts by DeepSeek and other Chinese AI labs to collect Claude outputs. Model distillation trains one model using another model’s outputs.
Chinese Foreign Ministry spokesperson Mao Ning said Tuesday that Beijing opposes expanding national security claims and politicizing technology and trade, and would defend Chinese enterprises’ legitimate rights. China was also reported last week to be considering export restrictions on 23 AI-related technologies, including chip designs and certain advanced AI models, citing security and misuse risks. The parallel restrictions underscore intensifying US-China competition in AI.
Positives
- Federal civilian agencies must remove existing DeepSeek-linked tools, addressing DHS concerns about sensitive government data.
- Third-party products using DeepSeek technology are included, extending protection beyond the startup’s own products and services.
- Agencies may request exceptions, preserving a route for specifically approved uses.
- DeepSeek’s low-cost, open-weight models have increased competition with OpenAI and Meta.
Risks & concerns
- Chinese law could make sensitive US government user data accessible to Chinese authorities, DHS warned.
- DeepSeek and five other Chinese AI companies face US accusations of extracting information from American systems through large-scale model distillation.
- OpenAI and Anthropic identified suspected or unauthorized efforts involving DeepSeek and outputs from their models, including Claude.
- China is considering restrictions on 23 AI technologies, potentially limiting access to chip designs and advanced models.
- The US order and China’s response deepen an already intensifying technology rivalry between the two countries.