A ministry official stated that South Korea’s industry ministry has temporarily limited employee access to the Chinese AI startup DeepSeek over security concerns. This decision follows a government advisory urging ministries and agencies to exercise caution when using AI services, such as DeepSeek and ChatGPT, in the workplace. Authorities plan to assess the potential risks associated with generative AI technology.
Acting President Choi Sang-mok described DeepSeek’s latest AI model, DeepSeek R1, as a “fresh shock,” highlighting its ability to deliver high performance at a significantly lower cost than existing models. Choi’s remarks came during a meeting with ministers, as the launch of DeepSeek R1 has led to widespread concern, especially within the public sector. The AI model’s emergence has caused turmoil in the stock market, with tech and semiconductor shares falling, and raised questions over data privacy, prompting the government to block access to official devices.
South Korea joins a growing list of nations blocking access to DeepSeek, the Chinese AI service, citing security concerns. Earlier this week, Australia and Taiwan took similar actions, banning DeepSeek from government systems due to the perceived national security risks posed by the technology. Australian intelligence officials raised alarms about the AI service, while Taiwan’s government also cited potential security threats.
The controversy surrounding DeepSeek extends beyond Asia, with Italy being the first country in Europe to take action. On January 28, Italy’s privacy watchdog mandated a nationwide block, questioning how the service handles user data. DeepSeek has been given 20 days to clarify its compliance with European data protection laws. Ireland has followed suit, requesting further details regarding its data policies from the company. In the United States, agencies like NASA and the U.S. Navy have restricted employees from using DeepSeek because they are concerned about privacy and security risks.
The surge in attention for DeepSeek came after the company’s mid-January release of its R1 AI model, which it claims matches or surpasses the performance of OpenAI’s popular GPT-3 model. The company’s ability to produce such advanced technology at a fraction of the cost compared to U.S. firms has caused a stir in the tech world. In December, DeepSeek revealed that its V3 large-language model was trained with under $6 million of computing power—much lower than the training costs for U.S. competitors, further fueling concerns about China’s rapid advancements in AI.
In response to these developments, South Korea is taking proactive steps to support its tech industry. Acting President Choi Sang-mok announced plans to establish a 34 trillion won ($23.4 billion) fund through the Korea Development Bank to foster innovation in cutting-edge technologies such as AI and semiconductors. Recognizing the competitive pressure, Choi emphasized the need for swift and comprehensive countermeasures.
The rapid advancement of DeepSeek’s AI technology has challenged the prevailing assumptions about the AI boom, particularly the need for massive investments in infrastructure like semiconductors and data centers. Investors who once believed that companies like OpenAI would maintain a significant lead in AI development are now questioning these assumptions. DeepSeek’s ability to produce high-performance AI models at a much lower cost has made it clear that China’s AI capabilities are advancing swiftly, disrupting the global tech landscape.
In response to the growing competition, South Korea’s government is moving to secure its position in the rapidly evolving tech sector. The government announced plans to create a 34 trillion-won ($23.4 billion) fund through the Korea Development Bank to boost investment in cutting-edge technologies, including AI and semiconductors.
Meanwhile, South Korean tech firms, such as SK Hynix and Naver, are becoming more cautious about using generative AI services. SK Hynix has restricted access to these services, allowing limited use only when absolutely necessary, while Naver has instructed its employees to avoid AI tools that store data outside the company’s network.
