menu

South Korea Accuses AI Chatbot ‘DeepSeek’ of Sharing User Data Without Permission

Did DeepSeek Send Your Data Abroad Without Asking?
April 25, 2025
South Korea Accuses AI Chatbot ‘DeepSeek’ of Sharing User Data Without Permission

South Korea’s data protection authority has accused Chinese AI chatbot DeepSeek of sharing user data with American and Chinese companies without getting proper permission. In an official statement on Thursday, the Personal Information Protection Commission (PIPC) claimed that DeepSeek transferred personal data and prompts to foreign companies without users’ clear consent. These transfers reportedly happened when DeepSeek launched its services in South Korea in January.

DeepSeek Sent Data to Chinese and American Firms Without Consent

According to the commission, DeepSeek shared sensitive data, including prompts, network details, and app-related information, with several companies, including Beijing Volcano Engine Technology, without following legal procedures. The South Korean government has now demanded the deletion of all AI prompt data sent to Volcano Engine and recommended the company establish clear legal guidelines before sending user information to foreign servers again.

DeepSeek Responds but Concerns Remain

In response, DeepSeek told South Korean authorities that it used Volcano Engine’s technology to improve security and user experience. However, the chatbot service later stopped sending any prompts from April 10 onwards. Despite this action, concerns remain high, especially due to DeepSeek’s Chinese ownership and the global suspicion surrounding data practices in China.

China Says It Does Not Force Companies to Misuse Data

China’s foreign ministry quickly responded to the news, stating that the Chinese government has never and will never instruct companies to collect or store user data illegally. This statement came as DeepSeek faced global criticism for data risks and possible government influence.

DeepSeek Became Popular Quickly—But Faced Global Scrutiny

Launched in January, DeepSeek’s R1 model gained global attention as a cheap and efficient alternative to ChatGPT. Many praised its performance, especially for being developed quickly at a low cost. However, experts also raised concerns about user privacy, Chinese policies, and the risk of data misuse. Several countries have restricted DeepSeek or banned