The United States government ordered federal civilian agencies to stop buying and using products from Chinese AI startup DeepSeek, according to TechCrunch. Announced by the Department of Homeland Security, the directive requires agencies to remove the technology from government IT systems due to national security risks tied to the company's operations in China.
Key takeaways
- Federal civilian agencies must remove DeepSeek AI from their systems.
- The order covers both direct products and third-party services built using the technology.
- China opposed the move through Foreign Ministry spokesperson Mao Ning.
- US and Chinese firms have previously raised model distillation concerns.
Why is the US banning DeepSeek AI?
TechCrunch reports that the Department of Homeland Security raised specific worries about how DeepSeek handles user data. Officials warned that sensitive information from American government users might be accessible to Chinese authorities under local laws. This new directive expands past a December 2025 law that limited the software's use in intelligence national security systems.
The restriction blocks new purchases and forces the removal of existing software. Agencies can still ask for exceptions if needed. This action arrives during intense technological competition between the US and China, where the Chinese startup competes with companies like OpenAI and Meta by offering affordable open-weight models.
How is China responding to the federal agency restrictions?
Beijing pushed back against the decision promptly. According to an official briefing cited by TechCrunch, Foreign Ministry spokesperson Mao Ning said Tuesday that Beijing opposes overstretching the concept of national security and politicizing technology and trade issues. The spokesperson added that China will resolutely safeguard the legitimate rights and interests of Chinese enterprises.
Meanwhile, Bloomberg reported last week that China is considering export restrictions on 23 AI-related technologies, including chip designs and advanced models. DeepSeek has faced global scrutiny since its R1 reasoning model debuted in January 2025, which caused a drop in Nvidia shares. Also, the US accused DeepSeek, Alibaba, Moonshot AI, and three other firms of using model distillation to extract information from American systems.
What it means for the tech industry
This mandate highlights the growing divide between American and Chinese technology ecosystems. OpenAI and Anthropic previously told lawmakers about suspected distillation attempts by Chinese labs looking to collect model outputs. As government regulations tighten, tech companies on both sides face stricter boundaries regarding data sharing, cross-border development, and advanced model training practices.
Frequently asked questions
The Department of Homeland Security cited national security concerns, warning that sensitive government user data could be accessible to Chinese authorities under Chinese law.
The order also covers third-party products and services built using DeepSeek technology, though federal agencies are permitted to request exceptions.
Foreign Ministry spokesperson Mao Ning stated that Beijing opposes politicizing technology and trade issues, adding that China will resolutely safeguard the legitimate rights of its enterprises.
A law passed in December 2025 previously targeted the use of DeepSeek in intelligence and national security systems before the latest directive extended rules to civilian agencies.
OpenAI and Anthropic raised concerns regarding suspected model distillation attempts, where Chinese labs allegedly collected outputs from American AI models to train their own systems.
Based on reporting by TechCrunch. Read the original article
