US Bans DeepSeek From Federal Civilian Agency Systems

The United States has ordered federal civilian agencies to remove DeepSeek artificial intelligence technology from their computer systems, escalating Washington's technology dispute with Beijing over national security and data protection. The Department of Homeland Security (DHS) disclosed the order on Monday, Oct. 19, 2026, while Chinese officials defended the rights of domestic technology companies.

Issued on July 24, the directive prohibits civilian agencies from purchasing DeepSeek products and services. It also covers third-party products and services developed using the Chinese company's AI models, according to the DHS notice published in the Federal Register.

The restrictions apply to agencies including the Treasury, State and Justice departments but exclude the Department of Defense. Federal executive agencies may request exceptions, including extensions to comply, by submitting written justifications to DHS within 60 days of the order's issuance.

The DHS identified potential exposure of sensitive government information to Chinese authorities and cybersecurity weaknesses as central reasons for the restrictions. The department warned that DeepSeek's models could provide access to information about American citizens and government programs if sensitive data were transferred to China.

The agency also raised concerns about possible malware and other security flaws, citing open-source research that identified multiple critical vulnerabilities in DeepSeek's iOS application. The DHS notice characterized the models as presenting multiple threat vectors.

DeepSeek is among China's best-known AI companies, and its open-source models are widely available internationally, including in the United States. The models are generally free to use, making the restrictions relevant not only to software supplied directly by the company but also to services built on its technology.

Chinese Foreign Ministry spokesperson Mao Ning responded to the US decision by defending Chinese businesses and criticizing Washington's approach to national security. Speaking at a regular press briefing in Beijing, she said China would protect its companies' legitimate interests.

China will resolutely safeguard the legitimate rights and interests of Chinese companies, Mao said.

She also objected to what Beijing considers the expansion of national security arguments into commercial and scientific matters.

China consistently opposes the overstretching of the concept of national security and the politicizing of science, technology, economic and trade issues, Mao said.

The latest directive follows earlier US government restrictions on DeepSeek. In March 2025, US Commerce Department bureaus informed employees that the application was prohibited on government devices, according to Reuters. In January 2025, the Defense Information Systems Agency moved to block access to DeepSeek models, according to previous media reports.

Congress subsequently restricted DeepSeek's use in intelligence community systems in December 2025. The newly disclosed DHS order expands the restrictions across federal civilian agencies while maintaining an exclusion for the Defense Department.

The dispute forms part of a broader competition between Washington and Beijing over advanced AI technology, computing infrastructure and national security. US-led export controls continue to restrict China's access to some of the world's most advanced AI chips and the machinery used to manufacture them.

Beijing introduced proposed restrictions on exports of certain advanced AI models the week before the latest reports, adding another point of tension between the two countries. Washington has also accused Chinese AI startups, including DeepSeek, of large-scale distillation, a process involving the extraction of capabilities from more advanced AI models developed by US companies. Beijing has rejected those allegations.

Despite the disagreements, the two governments have made some progress on AI governance cooperation, including a US proposal to establish a notification system for national security-related incidents. However, significant disputes over technology access, trade and security remain unresolved.

US President Donald Trump and Chinese President Xi Jinping are scheduled to meet in mid-November on the sidelines of the Asia-Pacific Economic Cooperation leaders' summit in Shenzhen, China. Artificial intelligence is expected to feature on the agenda.