DeepSeek barred from US federal civilian IT systems over security concerns

The US Department of Homeland Security has ordered all federal civilian agencies to remove products made by Chinese AI startup DeepSeek from government IT systems and stop buying them in the future. The department said the move was needed because of concerns about the security of the supply chain.

Washington: According to an official government notice, the US Department of Homeland Security (DHS) has ordered federal civilian agencies to remove products made by Chinese artificial intelligence (AI) startup DeepSeek from their IT systems. The agencies are also not allowed to buy these products, Yonhap news agency reported.

The order is explained in a notice that is scheduled to be published in the Federal Register on Tuesday. It covers IT products and services made by Hangzhou DeepSeek Artificial Intelligence Co. and its related companies.

The restrictions also cover products and services made by other companies that use DeepSeek technology.

The department said it issued the order under the Federal Acquisition Supply Chain Security Act of 2018. It did so after reviewing a recommendation from the Federal Acquisition Security Council and other related evidence.

The order applies to IT systems and purchases used by federal civilian agencies. It does not create a general ban on private businesses or individuals in the United States using DeepSeek products.

In February 2025, the South Korean government also took steps to limit the use of DeepSeek. Its foreign, trade and defence ministries blocked access to the Chinese AI service on government computers because of concerns about how it collected user data and the risk of information leaks.

Meanwhile, US cybersecurity company CrowdStrike said this week that a hacker whose identity is unknown, but who is believed to speak Chinese, used AI-powered hacking tools to break into several South Korean financial institutions and steal data.

In a report, CrowdStrike said the attacker used ARTEX, an open-source AI-powered tool for testing computer security that was developed in China. The attacker also used large language models (LLMs) to carry out cyberattacks between late September and early October.

These findings come as South Korean financial institutions face a series of data breaches. The affected institutions include Hana Bank, KB Kookmin Bank and Shinhan Bank. Financial authorities and investigators have started investigations into the incidents.