South Korea’s government has banned the use of the Chinese artificial intelligence model DeepSeek across key ministries, citing concerns over potential leaks of sensitive information. The move aims to safeguard confidential data during the use of generative AI services.
The Ministry of Foreign Affairs, Ministry of National Defense, and Ministry of Trade, Industry and Energy imposed restrictions on Feb. 5, according to government officials. The Defense Ministry said it had blocked access to DeepSeek on military internet-connected PCs, citing “numerous technical concerns.” Other ministries reportedly implemented similar restrictions on computers connected to external networks.
The ban follows a Feb. 4 directive from the Ministry of the Interior and Safety, which urged central government agencies and 17 provincial governments to exercise caution when using generative AI models such as DeepSeek and ChatGPT. The directive advised against inputting personal information and warned against fully trusting AI-generated results.
DeepSeek has faced criticism for its extensive data collection practices, including monitoring users’ keyboard input patterns. In response, Australia, Taiwan, and the U.S. state of Texas have banned its use on government devices. Italy has also blocked downloads of the platform from Google and Apple app stores.
South Korea’s decision marks a coordinated effort to address growing concerns over information security in the AI era.
The restrictions are now spreading to the private sector. Korea Hydro & Nuclear Power and KEPCO Plant Service & Engineering recently announced internal bans on DeepSeek. Tech companies such as Kakao and LG Uplus have also prohibited its use for work-related purposes. Major conglomerates, including Samsung, SK, and LG Electronics, have reportedly blocked access to external generative AI programs entirely.
