US Orders Federal Agencies to Purge DeepSeek AI From Government Systems

The U.S. Department of Homeland Security has ordered federal civilian agencies to stop buying DeepSeek products and to strip the Chinese AI startup's technology out of existing government IT systems.

The directive, issued over national security concerns tied to DeepSeek's operations in China, goes beyond the company's own products. It also covers third-party products and services built using DeepSeek technology, though agencies can request exceptions.

In short: if DeepSeek's models are anywhere in the stack, they have to go.

What does the order actually cover?

The DHS flagged two things: how DeepSeek handles user data, and what Chinese law could let Beijing see. The order warns that sensitive information from U.S. government users could potentially be accessible to Chinese authorities under Chinese law.

This is not Washington's first move against the company. A law passed in December 2025 already targeted DeepSeek's use in intelligence national security systems. The new DHS order extends those restrictions to federal civilian agencies, covering both new purchases and existing IT systems.

How did China respond?

Beijing pushed back the same week. Foreign Ministry spokesperson Mao Ning said Tuesday that China opposes "overstretching the concept of national security and politicizing technology and trade issues," according to an official briefing.

"We will resolutely safeguard the legitimate rights and interests of Chinese enterprises," Mao added.

The friction runs in both directions. China is reportedly considering export restrictions on 23 AI-related technologies, including chip designs and certain advanced AI models, to address security and misuse risks, Bloomberg reported last week.

Why has DeepSeek drawn this much scrutiny?

The Chinese AI lab has faced government and regulatory scrutiny worldwide since its R1 reasoning model shook the AI industry in January 2025, triggering a sharp sell-off in Nvidia shares.

The U.S. escalated the pressure in September, accusing 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.

Model distillation uses outputs from one AI model to train another. OpenAI and Anthropic have each raised concerns about DeepSeek's alleged use of the technique. Earlier this year, OpenAI told U.S. lawmakers it had observed suspected attempts by DeepSeek to distill its models, while Anthropic said it had identified unauthorized attempts by DeepSeek and other Chinese AI labs to collect outputs from Claude.

The backdrop is an intensifying AI race between the U.S. and China. DeepSeek challenges U.S. rivals, including OpenAI and Meta, with low-cost open-weight models.

The takeaway: the DHS order turns DeepSeek from a competitive threat into a prohibited vendor across the federal civilian government, and Beijing's contemplated export curbs on 23 AI technologies suggest the restrictions will likely travel in both directions.