South Korea Blocks Access to Chinese AI Startup DeepSeek Over Security Concerns

South Korea’s industry ministry has temporarily restricted employee access to Chinese AI startup DeepSeek due to security concerns, joining a growing list of governments taking action against the company. On Tuesday, the government issued a cautionary notice advising ministries and agencies to be wary of AI services, including DeepSeek and ChatGPT.

State-run Korea Hydro & Nuclear Power blocked DeepSeek earlier this month, while the defense ministry has restricted access on military-use computers. The foreign ministry has also limited access on computers connected to external networks, according to Yonhap News Agency.

DeepSeek has yet to comment on the restrictions. While it remains unclear if South Korea has imposed similar measures on ChatGPT, the move aligns with global trends. Australia and Taiwan recently banned DeepSeek from government devices over security risks. In January, Italy’s data watchdog ordered the startup to block its chatbot over privacy concerns. Governments in the U.S., India, and Europe are also reviewing the implications of using DeepSeek.

South Korea’s privacy watchdog plans to investigate how DeepSeek manages user data. Meanwhile, local tech firms are tightening AI policies. Kakao Corp advised employees to avoid DeepSeek following its partnership with OpenAI. SK Hynix, a leading AI chipmaker, has restricted generative AI use, permitting access only when necessary. Naver, South Korea’s top web portal, has urged employees to avoid AI services that store external data.

DeepSeek made waves last month by launching AI models it claims rival U.S. competitors at a lower cost. However, rising security concerns are prompting governments and companies worldwide to reassess their AI strategies.