While we have seen numerous instances of the AI model from China, DeepSeek, making significant strides both globally and in the domestic market, there are also areas where it appears to be struggling. One such case comes from South Korea, where the popular AI model has been temporarily suspended by the government due to “privacy-related shortcomings.”

The service is expected to resume once improvements are made in accordance with the Domestic Personal Information Protection Act (Protection Act). This isn't the first time DeepSeek has faced challenges in South Korea, as reports indicate that even before the government's announcement, some ministries, agencies, and companies had already banned its use.

DeepSeek neglected to consider domestic protection laws

DeepSeek currently operates in two forms; one is a chatbot service and enterprise API directly operated by DeepSeek, and the other one is an open-source ‘DeepSeek-R1’ model developed by DeepSeek that anyone can download and install on their own server to use. Among these, the chatbot service and the enterprise API were pointed out as problematic, as DeepSeek collects extensive user personal information and the data is stored on Chinese servers.

Choi Jang-Hyeok, vice chairman of the Personal Information Protection Commission previously sent an official inquiry to the DeepSeek headquarters on January 31 regarding the collection and processing of personal information and began conducting their own analysis of the service. And this is what he had to say;

“As a result of our own analysis, we confirmed some insufficiencies in the communication functions with third-party businesses and in the personal information processing policy that were pointed out by domestic and foreign media outlets.”

Choi Jang-Hyeok, VC of Protection Act commission. Sources: zdnet (translated)

Is your data safe with DeepSeek?

The temporary suspension of DeepSeek in South Korea over privacy concerns might put the AI model in the global spotlight, raising questions about whether our data is truly safe with DeepSeek. While DeepSeek has significantly pushed other AI companies, like OpenAI, to accelerate their efforts, issues related to user information and personal data could turn this rapid rise into a downward spiral.

That said, the government has not provided an exact date for DeepSeek's return but has stated that it will inevitably take a considerable amount of time to bring the service in line with domestic laws. However, existing app and web service users will still be able to access DeepSeek, though the government advises caution.

“Last year, the preliminary inspection of major AI services such as OpenAI, Google, and MS (a total of six services) took about five months, but this inspection is limited to one operator, and it is expected to proceed more quickly due to the accumulated experience and know-how so far.”

Choi Jang-Hyeok (translated)

About the Author

Hassam Nasir

Hassam boasts over seven years of professional experience as a dedicated PC hardware reviewer and writer.