US bans DeepSeek AI products from federal agencies
Washington is afraid that DeepSeek is spilling the beans to Beijing.
Image by Cybernews.
Follow us
- The US government banned DeepSeek products and services across federal agencies, with two limited exceptions.
- Agencies must remove DeepSeek tools immediately and file exception requests within 60 days.
- Officials warn DeepSeek could expose sensitive prompts, user data, and device fingerprints to Chinese authorities.
- The US cites reported app vulnerabilities and follows similar restrictions in Europe, Taiwan, India, and Canada.
Key Takeaways by nexos.ai, reviewed by Cybernews staff.
The United States government is banning the use of all products and services developed by the Chinese AI company DeepSeek.
The Department of Homeland Security has published a notice that was issued by the Secretary of Homeland Security on July 26th, 2026.
The order says that all government agencies are prohibited from using products and services developed by DeepSeek, with the exception of the Department of War and the Office of the Director of National Intelligence.
In addition, all government agencies are ordered to remove all DeepSeek and DeepSeek-affiliated products immediately. Requests for exceptions must be submitted within 60 days and will be reviewed by the Cybersecurity and Infrastructure Security Agency (CISA).
The notice states that information and communications technology are vital to the US government systems. DeepSeek, a privately-owned AI company based in the People’s Republic of China (PRC), presents “multiple threat vectors,” the Department of Homeland Security argues.
The US government is concerned that sensitive information could fall into the hands of the Chinese government.
“DeepSeek servers are located in the PRC, and the PRC legal system permits extensive authority to government cyber and intelligence functions to access private company infrastructure. For DeepSeek-hosted instances of their models, DeepSeek’s account generation process collects personally identifiable information (PPI) and may also track and digitally fingerprint users,” the notice says.
Like other AI models, DeepSeek features a prompt-and-response process. The US government fears that DeepSeek is sharing sensitive prompts from government officials and government contractors with the Chinese government.
Furthermore, the government claims that DeepSeek’s AI models may contain “malware and other security vulnerabilities.” According to security researchers, DeepSeek’s iOS mobile app contains multiple critical vulnerabilities.
The US isn’t the first country to ban DeepSeek. Belgium, Italy, and Taiwan have also prohibited the use of DeepSeek by government officials because of potential security risks. Canada, the Czech Republic, India, and the Netherlands have imposed restrictions on the use of DeepSeek by civil servants.
