The US government has drawn its hardest line yet against Chinese AI. The Department of Homeland Security has ordered federal civilian agencies to remove DeepSeek's artificial intelligence products — and any third-party services built on its models — from their IT systems, according to a federal notice published on October 19, 2026 and reported by Reuters.
What the Order Does
- Removal: Federal civilian agencies must strip all products developed by Hangzhou DeepSeek Artificial Intelligence Co. and its affiliates from their information systems — including third-party products and services developed using DeepSeek implementations.
- Purchase ban: Agencies are barred from acquiring DeepSeek products and services going forward.
- Legal basis: Issued under the Federal Acquisition Supply Chain Security Act of 2018, following a review by the Federal Acquisition Security Council.
- Order date: Signed by Homeland Security Secretary Markwayne Mullin on July 24, 2026; filed for public inspection on October 19 and scheduled for Federal Register publication on October 20, 2026 (FR Doc. 2026-21447).
Why Washington Calls It a Risk
The DHS notice says DeepSeek models present "multiple threat vectors" for federal systems. The department warned that use of the products "may result in the transfer of sensitive data to the PRC if used by US government employees or vendors," adding that the models "themselves may contain malware or other security vulnerabilities."
According to reporting on the notice, DHS also warned that prompts entered into DeepSeek applications could send sensitive data to China, where it could be combined with public records to build "actionable intelligence" on US persons — and that the technology could be used to "track or digitally fingerprint" users.
The Backstory: From R1 Shock to Federal Ban
DeepSeek shook up global markets and the AI industry in 2025 with the release of its R1 model — trained, it said, using less powerful chips and at a fraction of the cost of American rivals, yet outperforming models from OpenAI and Anthropic on several complex tests. The release sent major US tech stocks into a stunning $1 trillion rout and was called a "wake-up call" by the US president at the time.
Scrutiny followed fast:
- The House Select Committee on China published a report in April 2025 claiming DeepSeek represented a "profound threat to our nation's security."
- In June 2025, US lawmakers introduced the "No Adversarial AI Act" seeking to restrict federal use of China-linked AI tools such as DeepSeek.
- South Korea's foreign, trade and defense ministries blocked DeepSeek on government computers in February 2025 over data-collection concerns.
- Some US federal agencies, including the Commerce Department and the Navy, had already blocked DeepSeek on work devices before this order.
What It Does NOT Do
Importantly, the order applies only to federal civilian agency IT systems and procurements. It is not a general ban on DeepSeek for private businesses or individuals in the United States. Civilian executive branch agencies are covered; national security systems are handled separately.
What Happens Next
Agencies now face the task of auditing their systems for DeepSeek-derived code — a tricky job, since the restriction also covers third-party services developed using DeepSeek implementations, which may not advertise their AI's origins. The notice allows agencies to request exceptions in writing with compelling justification and proposed mitigations.
The Chinese embassy in Washington did not immediately respond to requests for comment on the order. For the broader AI industry, the move is another sign of the hardening split between the US and Chinese technology ecosystems — and a preview of the compliance choices governments and companies worldwide will increasingly face.
Disclaimer: Regulatory details are as reported; agency implementation timelines may vary. Sources: Reuters, South China Morning Post (October 19, 2026).


