DeepSeek and Privacy: Why Korea Paused This Chinese AI Disruptor

By Chyung Eun-ju and Joel Cho

When DeepSeek sprang onto the global AI scene earlier this year, it caused waves not just in the tech world but throughout the global market. Developed by a Chinese startup led by Liang Wenfeng, this chatbot quickly gained a large international following, topping app store rankings and being compared to OpenAI's ChatGPT.

However, this rapid rise brought with it significant concerns regarding data privacy, security, and regulatory compliance, similar to challenges faced by other AI technologies. Recently, South Korea joined the list of countries scrutinizing DeepSeek, temporarily halting new downloads of the application due to worries about its handling of user data.

DeepSeek's success is largely attributed to its ability to provide high-performance AI at a price lower than its Western counterparts. However, as users become increasingly aware of the importance of data protection for maintaining their privacy, questions surrounding DeepSeek's data practices have become more prominent.

Reports have emerged indicating that DeepSeek shares data with ByteDance, the company behind TikTok. This has raised fears of user data potentially being transferred to entities linked with the Chinese government. The implications of China's strict data protection laws have led regulators to doubt whether DeepSeek is capable of adequately safeguarding user privacy, prompting Korea's Personal Information Protection Commission (PIPC) to act. As a result, the PIPC decided to temporarily remove DeepSeek from app stores in the country while it investigates the company's data handling practices. While those already using the service can still access it, new users are unable to download the app. This action aligns with similar measures taken by other nations such as Italy, Australia, and Taiwan, which have all limited or banned DeepSeek to varying degrees.

Concerns regarding data security and transparency have compelled several countries to restrict DeepSeek. In Italy, the Data Protection Authority placed limits on DeepSeek’s user data processing, citing a lack of transparency. Taiwan prohibited government agencies from utilizing DeepSeek services, and Australia outright banned the tool on all government devices, referencing national security issues. The argument from these governments is that DeepSeek's vague data handling practices present a potential risk to privacy, especially given China's National Intelligence Law, which allows the Chinese government access to corporate data. Compared to Western AI companies that typically face regulatory challenges but are often still allowed to operate, Chinese firms like DeepSeek are frequently viewed as higher security risks, leading to stronger restrictions globally.

Korea’s data protection laws, similar to the European Union’s General Data Protection Regulation (GDPR), focus on principles such as data minimization, purpose limitation, and transparency. If DeepSeek is found to violate any of these principles, it could face serious regulatory consequences. For a nation like Korea, recognized as a leader in technological advancement with robust data protection standards, DeepSeek's situation serves as a crucial test of its regulatory system.

On February 17, the PIPC issued a press release stating, "DeepSeek has temporarily suspended its service in Korea as of February 15, 2025, to enhance compliance with the Personal Information Protection Act (PIPA). The company plans to resume service once necessary improvements are implemented."

DeepSeek’s rise also showcases China's ambitions to lead the AI sector independent of Western influence. However, its swift foray into global markets has brought back worries regarding data sovereignty. The Chinese National Intelligence Law gives the state authority to compel companies and citizens to assist with national defense and intelligence operations, further complicating the data privacy landscape.

This suspension has captured the interest of prominent Korean AI companies like Naver and Kakao. In response, Naver is working on improving its AI model, HyperCLOVA X, to enhance its efficiency and decrease operational costs in order to compete with DeepSeek's low-cost, high-performance strategy. Despite its focus on developing internal AI, Naver remains open to collaboration with external technology providers.

Kakao has opted to partner with OpenAI, incorporating various AI models into its offerings rather than relying solely on in-house development. Nonetheless, Kakao's management has expressed apprehensions regarding DeepSeek's security and trustworthiness, hinting that increased investments in safety protocols might negate some of the cost advantages.

Korea has started to establish itself as an "AI middle power," alongside countries like Japan, the European Union, and nations in the Middle East. The Boston Consulting Group (BCG) notes that as competition between the U.S. and China escalates, mid-tier AI nations such as Korea have a unique opportunity to enhance their global influence.

A significant advantage for Korea lies in its leadership in AI-related semiconductor manufacturing, with companies such as Samsung and SK hynix dominating the global memory chip market. However, there are growing concerns that this edge may not be sustainable, as countries like the U.S. and EU are heavily investing in domestic semiconductor production to minimize reliance on foreign supply chains.

As Korea navigates developments surrounding DeepSeek, its case symbolizes the ongoing discussions regarding AI governance, data privacy, and the balance between innovation and regulation. While the future of DeepSeek's operations in Korea remains uncertain, its influence on AI policy and industry competition is unmistakably significant.

Chyung Eun-ju is a tech research associate at Donghyun ASP, with a bachelor's in business and a master's in marketing from Seoul National University. Joel Cho is a lawyer specializing in IP and digital law.

AI, Privacy, Korea