
The United States Department of Homeland Security (DHS) has ordered the removal of all artificial intelligence technology developed by the Chinese firm DeepSeek from most civilian government systems. The directive mandates that agencies, including the Departments of Treasury, State, and Justice, purge the company’s software from their networks. While the public announcement of the security-related ban occurred in October 2025, the order has been in effect since late July, reflecting intensified efforts by the US administration to mitigate perceived risks from Chinese-owned technology.
The ban specifically targets civilian agencies, notably excluding defense organizations from this particular directive. DeepSeek has gained significant international traction for its open-source AI models, which are often cited as being substantially more cost-effective to develop and operate than those of its American competitors. This popularity has led to the integration of DeepSeek’s technology into various third-party offerings, including some used by prominent US tech companies. Despite this widespread adoption, the DHS and the White House have raised alarms regarding the security implications of relying on AI frameworks developed by firms subject to Chinese government influence.
While the DHS has not released specific technical details regarding the vulnerabilities or intelligence that prompted the move, the action follows a pattern of heightened scrutiny of Chinese AI developers. US officials have frequently expressed apprehension regarding data privacy and the potential for foreign interference through software embedded in critical government infrastructure. For DeepSeek, the ban represents a significant hurdle in the American market, where its low-cost model had previously been seen as a competitive advantage. The move underscores the deepening technological divide between Washington and Beijing as both nations vie for dominance in the artificial intelligence sector.