THE U.S. Department of Homeland Security has ordered federal civilian agencies to remove DeepSeek’s artificial intelligence products and any third‑party tools that were developed using DeepSeek from their IT systems, with the notice dating back to late July.
The Federal Register notice argues that DeepSeek’s models pose multiple threat vectors, noting that the company’s servers are located in China and raising concerns that Chinese authorities could potentially access private U.S. company infrastructure and sensitive data. DeepSeek’s open‑source models are widely available, including in the United States, where they are largely free to use.
The action is part of broader U.S. restrictions on access to certain advanced AI capabilities and chips, though the government’s public statements emphasise national security considerations rather than a conclusion of specific exploits.
Beijing has pushed back against the move, with Foreign Ministry spokesperson Mao Ning asserting that China will safeguard the legitimate rights of Chinese firms and opposing what it calls the overextension of national security concepts. China’s commerce ministry followed with a strongly worded statement claiming the U.S. is discriminating under the guise of security and that the decision undermines broader China–U.S. economic and trade relations.
They pledged to monitor developments and defend Chinese interests, while DeepSeek did not immediately respond to requests for comment. The article notes continued but uneasy engagement between the two powers on AI governance, and mentions that AI policy is expected to feature on the agenda when Trump and Xi meet at the APEC summit in November.