The U.S. Department of Homeland Security has ordered federal civilian agencies to purge DeepSeek products from their IT systems, escalating Washington’s campaign against the Chinese AI company over national security concerns.
Agencies must immediately exclude covered products from procurement and remove existing installations from agency systems, or seek an exception. Exception requests must be filed in writing within 60 days of the order’s issuance, explaining the justification and proposed risk mitigations. CISA will review each request and issue written approvals or denials.
DHS said the action stems from fears that the Chinese government could access DeepSeek’s infrastructure, collect personal information, and transfer sensitive government data to China. The notice cited DeepSeek’s China-based servers and flagged potential malware and security flaws, pointing to a February 2025 NowSecure report that found unencrypted data transmissions and data sharing with ByteDance in the DeepSeek iOS app.
The ban reaches beyond the chatbot and API to any tool built using DeepSeek implementations, making compliance a software supply chain exercise. It applies to federal civilian agencies — including the Treasury, State, and Justice departments — but not to private consumers, state and local governments, or Defense Department national security systems.
The notice was published in the Federal Register on October 20, though the underlying order was issued in late July — making this a public disclosure of an earlier decision, and the most forceful U.S. federal action yet against a frontier model from China. For agencies that relied on the inexpensive open-weight models, the clock is ticking: pivot to alternatives like ChatGPT and Claude — or make a compelling case to CISA. See our best AI chatbots compared.
Source: BleepingComputer, TechSpot
