US DHS Orders Federal Civilian Agencies to Remove DeepSeek and Derived Services

According to TechCrunch, the U.S. Department of Homeland Security (DHS) has issued an order requiring all federal civilian agencies to cease purchasing products and services from Chinese AI startup DeepSeek and to remove them from government IT systems. This move marks a significant expansion of restrictions on DeepSeek, extending beyond previous limitations that targeted only intelligence community national security systems.

The scope of the order goes beyond DeepSeek's own chatbot or API; it explicitly covers "third-party products and services built using DeepSeek technology." This means software vendors integrating DeepSeek's open-weight models into their offerings face exclusion from government procurement. Although agencies can request exceptions, this regulation significantly increases compliance costs and forces middleware and vertical application developers who rely on DeepSeek's low-cost inference capabilities to reassess their strategies for the government market.

The primary concern cited by DHS involves data handling. The order warns that sensitive information generated by U.S. government users through DeepSeek could potentially be accessible to Chinese authorities under Chinese law. This rationale aligns with recent accusations from U.S. labs like OpenAI and Anthropic regarding DeepSeek's alleged large-scale model distillation, further intensifying geopolitical tensions in the AI sector between the U.S. and China.

In response, Chinese Foreign Ministry spokesperson Mao Ning stated opposition to "overstretching the concept of national security and politicizing technology and trade issues," pledging to safeguard the legitimate rights of Chinese enterprises. Meanwhile, reports indicate that China is considering export controls on 23 AI-related technologies, including chip designs and advanced models, reflecting an adversarial escalation in regulatory measures.