The Department of Homeland Security has ordered DeepSeek, the Chinese artificial intelligence company, out of civilian executive branch agencies’ technology systems. Officials cited fears that Beijing could tap sensitive American data to “produce actionable intelligence.” Beijing didn’t take it quietly. China accused Washington of discrimination “under the pretext of so-called national security.”
Issued in July, the order took effect this summer and appeared publicly on Monday. It covers more than DeepSeek’s own products: it also reaches products from other firms that rely on the company’s open-source offerings, and bars agencies from acquiring any of them in the future.
The Security Rationale
The concern at the heart of the order is structural, not technical. U.S. officials said China’s legal system allows its government extensive access to Chinese company infrastructure. They argue that arrangement leaves the U.S. government vulnerable to exposure of sensitive information.
Put another way, the question isn’t whether a particular chatbot behaves well on a given day. It’s whether a company operating under that legal regime can credibly promise to keep a foreign government’s hands off the data moving through its systems. Homeland Security concluded it can’t, at least not to a standard federal agencies should accept.
The order’s breadth follows that logic. It reaches products built on DeepSeek’s open-source offerings, closing a back door a narrower ban might have left open.
Beijing’s Response
Chinese officials called the ban a pretext. They accused the United States of using disingenuous national security concerns to suppress Chinese companies.
China’s Ministry of Commerce said in a statement that China had passed laws to protect cybersecurity and data privacy and ensure companies can comply with relevant laws worldwide. It called on the United States to “withdraw” restrictive measures.
Mao Ning, a spokeswoman for China’s foreign ministry, pressed the argument further at a briefing on Tuesday. “China has consistently opposed the overgeneralization of the concept of national security and the politicization of science, technology and trade issues,” she said. “We will resolutely safeguard the legitimate rights and interests of Chinese enterprises.”
The Chinese statements say nothing about the specific point U.S. officials raised: Chinese law gives the government broad access to company infrastructure. Assurances about privacy legislation sit alongside that access. They don’t answer it.
The Larger Contest
The dispute comes amid fierce competition for A.I. primacy between the two countries. Some experts have likened the race to the 20th-century nuclear arms race, with potentially huge stakes for humanity. Whatever one makes of that comparison, the stakes help explain why Washington would rather lose access to a cheap tool than risk a data breach at the source.
The ban has limits. It applies to civilian executive branch agencies’ own technology systems. It doesn’t dictate what private companies or individual citizens may use, and leaves the market to sort out the rest.
Why This Matters:
The order draws a line on a question every government will face: how much trust to extend to technology built under a legal system that lets the state reach into private companies. For Washington, the answer is that sensitive federal data shouldn’t sit on infrastructure Beijing can access. The approach targets government systems rather than the broader economy, leaving decisions about private adoption to businesses and consumers. But the stakes are rising. With the two powers locked in a contest over A.I. leadership, expect more scrutiny of supply chains, open-source dependencies, and who ultimately controls the data. Procurement rules, not slogans, may end up deciding how much of that contest Washington wins.
- deepseek
- china
- united states
- department of homeland security
- artificial intelligence
- government systems
- national security
- us-china tensions