US Homeland Security Bans DeepSeek AI From Civilian Federal Agencies Over Data Concerns

The US Department of Homeland Security is banning civilian federal agencies from using Chinese AI model DeepSeek and its affiliates. A notice due to appear in the Federal Register cites the Federal Acquisition Supply Chain Act and a review by the Federal Acquisition Security Council. It bars agencies from buying IT products and services from DeepSeek's parent company, Hangzhou DeepSeek Artificial Intelligence. A letter from Secretary Markwayne Mullin, dated June 26, ordered the exclusion, which suggests the ban has been in place for months.

Mullin said DeepSeek's models pose "multiple threat vectors" to federal IT systems. Some risks relate to the company's country of origin, while others are privacy and mass data collection concerns that could also apply to other Chinese and US AI models. He said DeepSeek models store data on servers in China, where national security laws give the government wide latitude to request data from domestic companies. Models hosted directly by DeepSeek collect personally identifiable information and trace users' "digital fingerprints" online. Like other major AI models, DeepSeek relies on a prompt and response process, which Mullin said may result in sensitive data reaching the Chinese government.

Mullin warned that this information, alone or combined with public data, could produce actionable intelligence on US persons and government programs. The decision follows accusations, made more than a year ago by US frontier AI companies and the federal government, that DeepSeek's makers "systematically" distilled other firms' models. CyberScoop has asked CISA and DHS for further details.