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 Chinese AI startup DeepSeek's products from government IT systems and halt future procurement, citing supply chain security concerns.

Washington: The US Department of Homeland Security (DHS) has ordered federal civilian agencies to remove products developed by Chinese artificial intelligence (AI) startup DeepSeek from their information technology systems and barred them from procuring such products, according to a federal notice.

The order, detailed in a notice scheduled to be published in the Federal Register on Tuesday, applies to information technology products and services developed by Hangzhou DeepSeek Artificial Intelligence Co. and its affiliated entities, reports Yonhap news agency.

---

The restrictions also cover third-party products and services developed using DeepSeek implementations.

The department said the order was issued under the Federal Acquisition Supply Chain Security Act of 2018 following a review of a recommendation by the Federal Acquisition Security Council and related evidence.

The order applies to federal civilian agency IT systems and procurements, rather than imposing a general ban on the use of DeepSeek products by private businesses or individuals in the United States.

The South Korean government also moved to restrict the use of DeepSeek in February 2025, with its foreign, trade and defence ministries blocking access to the Chinese AI service on government computers amid concerns over user data collection and potential information leaks.

Meanwhile, US cybersecurity firm CrowdStrike said this week that an unidentified hacker believed to be a Chinese speaker used artificial intelligence (AI)-powered hacking tools to breach multiple South Korean financial institutions and steal data.

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

The findings come amid a series of data breaches at South Korean financial institutions, including Hana Bank, KB Kookmin Bank and Shinhan Bank, prompting financial authorities and investigators to launch probes into the incidents.