The US Department of Defense has swiftly intervened to prevent access to DeepSeek, an AI firm based in China, after it was revealed that employees at the Pentagon had linked their work computers to Chinese servers while utilising the AI service.
DeepSeek’s user agreement explicitly indicates that data is stored on servers located in China and is subject to Chinese law, which requires compliance with the nation’s intelligence agencies. Despite these clear stipulations, several Pentagon staff members reportedly engaged with the service for a minimum of two days, which raised security alarms.
On January 28, the Pentagon’s Defense Information Systems Agency (DISA), responsible for managing military IT networks, initiated measures to block access to DeepSeek across certain systems. Nevertheless, reports indicate that some employees continued to access the AI tool even after this action was taken.
The official justification for the blockade was cited as “operational reasons,” but this decision followed heightened concerns regarding the potential security threats associated with using DeepSeek, particularly due to its ties to China.
Military Concerns Regarding Data Security and Privacy
DeepSeek’s privacy policy makes it clear that user data is retained in China and governed by Chinese legislation, which has alarmed officials within the Pentagon. Although the company asserts that it adheres to standard privacy protocols, the association with Chinese servers has raised critical issues surrounding data security, privacy, and possible espionage threats.
According to Bloomberg, military personnel had begun downloading earlier versions of DeepSeek’s software onto their workstations in the fall of 2024 without initially recognising the implications of connecting to Chinese servers.
The surge of interest in DeepSeek’s latest release has led various military branches to take action in locating and removing any code from Chinese-origin chatbots present on employees’ devices.
Different military branches are responding to DeepSeek usage among employees with varying approaches. The Navy recently issued a ban on any use of DeepSeek due to potential security and ethical concerns related to its origins and applications.
Meanwhile, the Department of the Air Force lacks specific regulations regarding DeepSeek but already prohibits the use of sensitive public information in commercial generative AI systems without appropriate approvals. The Army had previously issued guidance in June 2024 highlighting “unique challenges regarding data privacy, security, and control over generated content,” urging commands to establish suitable governance processes while avoiding outright bans on generative AI tools.
Some service members are now contemplating new policies that would explicitly prohibit the use of generative AI models from China.