The U.S. Department of Homeland Security (DHS) has ordered all federal civilian agencies to remove products developed by Chinese artificial intelligence (AI) startup DeepSeek from their information technology (IT) systems and barred them from procuring such products. This directive mandates that agencies purge existing tools and applications from their servers, effectively severing the operational link between the U.S. civilian government and the company.
The timing of this directive follows a broader context of the U.S. government’s stance on DeepSeek. The progression suggests a tightening of security posture, moving from a general stance to a targeted operational directive that mandates the removal of the technology from federal infrastructure.
The DHS’s decision rests on a fundamental architectural concern: DeepSeek’s servers are located in China. In the world of AI, where data is processed matters. The rationale provided for this ban centers on the risk of data transfer to China. If the AI model is running on servers within Chinese territory, the data that passes through those systems is subject to the laws and jurisdiction of the People’s Republic of China.
Korea’s move was driven by similar fears of data collection and potential information leaks.
Complicating this risk is the broader legal landscape. The ban cites Chinese legislation that allows for state access to private infrastructure. The U.S. government fears that sensitive civilian data could be accessed by the Chinese state if it were processed on DeepSeek’s servers.
The South Korean government has also restricted the use of DeepSeek. Korea’s move was driven by similar fears of data collection and potential information leaks. This international concurrence in restricting the startup adds weight to the security concerns recognized by allied governments.
Security forces are taking a hard line against integrating foreign AI into civilian IT systems due to the risk of espionage. A cybersecurity firm reported an AI-powered breach by a Chinese speaker using ARTEX, an AI hacking tool. The DHS’s ban is a protective measure to ensure that federal systems are not exposed to vulnerabilities created by the integration of technology subject to foreign jurisdiction.