The United States has ordered federal civilian government agencies to remove artificial intelligence products developed by Chinese startup DeepSeek from their information technology systems, citing potential national security and data security risks.
The US Department of Homeland Security (DHS) issued the order, which also prohibits federal civilian agencies from purchasing DeepSeek products and services. The restrictions extend to products and services developed by affiliated entities and third parties using DeepSeek technology.
The order was issued under the Federal Acquisition Supply Chain Security Act of 2018, following a review of a recommendation and supporting evidence from the Federal Acquisition Security Council. The Department of Homeland Security’s notice was published in the Federal Register on October 20, 2026.
The Department of Homeland Security identified several potential security risks associated with DeepSeek’s technology. According to the department, using the products could result in sensitive information being transferred to the People’s Republic of China when accessed by US government employees or contractors.
The department also warned that the AI models could potentially generate actionable intelligence about US individuals or government agencies. It further raised concerns that the models themselves could contain malware or other security vulnerabilities.
The order forms part of the US government’s efforts to protect federal information technology systems and supply chains from potential foreign security threats.
Under the directive, agencies must remove DeepSeek and affiliated products from their IT systems and refrain from acquiring covered products and services. The restrictions also apply to third-party products and services developed using DeepSeek implementations, extending the scope beyond the company’s own applications.
However, the order does not impose a blanket prohibition on DeepSeek for private individuals or businesses across the United States. Its stated scope concerns federal civilian agency systems and procurement.
China criticised the restrictions on Tuesday, October 20, arguing that the decision unfairly targets Chinese businesses and harms their interests.
According to Reuters, China’s Ministry of Commerce called on Washington to revoke the measures, describing them as discriminatory. Beijing also argued that the decision disregarded understandings reached between US President Donald Trump and Chinese President Xi Jinping.
China said it would take necessary measures to protect the legitimate interests of its companies. The dispute comes amid continuing tensions between Washington and Beijing over technology, trade and national security.
DeepSeek gained international attention in January 2025 following the release of its R1 reasoning model. The company attracted interest for developing AI models that competed with established systems while drawing attention to the cost of building advanced artificial intelligence.
Its models have also contributed to wider discussions about open-weight AI systems, which allow developers to access model parameters and, subject to the applicable licences, adapt and deploy them.
The company’s rapid rise has prompted scrutiny from governments and technology companies over data handling, security and competition in the AI industry.
The latest US order expands restrictions on DeepSeek within the federal government. Several US agencies, including the Department of Commerce and the US Navy, had previously restricted the AI service on government devices. Some US states, including New York, Tennessee and Virginia, have also introduced restrictions on its use in government systems.
The decision highlights the growing importance of cybersecurity assessments in government technology procurement as artificial intelligence adoption expands. It also underscores the continuing competition between the United States and China over AI development, access to technology and the protection of sensitive information.
