Against Deepseek
South Korea’s government ministries and police have taken decisive action against Deepseek, a Chinese artificial intelligence startup, by blocking its access to work computers after the company failed to respond to a formal data request regarding how it manages user information. The move highlights growing concerns about data security and privacy in the context of generative AI services. On Thursday, South Korea’s police confirmed that they had blocked access to Deepseek’s services, while the Ministry of Trade stated that access had been temporarily restricted on all PCs under its jurisdiction. This decision underscores the seriousness with which South Korean authorities are treating potential risks associated with foreign AI platforms.
Deepseek gained attention last month when it launched its R1 Chatbot, touting it as a groundbreaking innovation capable of matching the performance of leading artificial intelligence systems in the United States at a fraction of the cost. The company claimed that its technology could revolutionize the global AI industry by offering advanced capabilities without requiring massive investments. However, this bold claim has drawn scrutiny from governments worldwide, including South Korea, France, and Italy, all of whom have raised questions about the company’s data-handling practices. These countries have sought detailed explanations from Deepseek regarding how it collects, processes, and protects user information.
In South Korea, the issue came to a head when Deepseek did not respond to an inquiry from the Personal Information Protection Commission (PIPC), a regulatory body tasked with safeguarding personal data. In response, multiple government ministries took immediate steps to limit access to Deepseek’s services to prevent any potential leaks of sensitive information through its generative AI tools. For instance, the Ministry of Defense announced that it had implemented specific measures to block Deepseek’s access to military-related PCs connected to the internet. An official from the ministry emphasized that these actions were necessary given the heightened security risks posed by the unauthorized use of AI technologies within defense networks.
Deepseek international pushback
The Ministry of Defense, which oversees active soldiers stationed near North Korea—a region fraught with geopolitical tensions—has also reiterated strict guidelines for using generative AI tools. These guidelines address both security and technical concerns, ensuring that units and personnel adhere to established protocols to mitigate risks. Similarly, the Ministry of Trade acted swiftly when Deepseek ignored the PIPC’s inquiry, stating that restrictions were imposed due to the company’s non-compliance. Meanwhile, the Ministry of Finance confirmed that it had implemented measures prohibiting the leakage of personal and confidential information to Deepseek across all employees’ devices.
This is not the first time Deepseek has faced international pushback over its AI offerings. Last week, Italy launched an investigation into the R1 model and blocked the platform from processing data belonging to Italian users. Australia followed suit, banning Deepseek from all government equipment based on recommendations from its security agencies. Such coordinated efforts reflect a broader trend among nations to scrutinize foreign AI companies operating within their borders, particularly those hailing from China, where concerns about state influence and data sovereignty persist.
Deepseek under U.S. export control
Deepseek has defended itself by highlighting its reliance on H800 chips, a type of semiconductor permitted for sale under U.S. export control regulations as of 2023. These chips power the company’s large training models, enabling it to deliver robust AI solutions despite hardware limitations. However, this explanation has done little to assuage fears among regulators who remain focused on ensuring transparency and accountability in data management practices.
South Korea’s stance on this matter carries significant weight, especially considering its position as a hub for cutting-edge semiconductor technology. Local chip giants like Samsung Electronics and SK Hynix supply some of the most advanced chips used in AI servers globally. As such, South Korea plays a pivotal role in shaping the future of AI development and deployment. By taking a firm stand against Deepseek, the country is signaling its commitment to protecting national interests while fostering responsible innovation in the AI sector.
In conclusion, the blockade of Deepseek by South Korean authorities serves as a cautionary tale for AI startups seeking to expand internationally. It underscores the importance of adhering to local regulations and maintaining open communication with regulators. At the same time, it highlights the delicate balance between embracing technological advancements and safeguarding sensitive information. As more countries adopt similar measures, the global AI landscape will likely see increased emphasis on compliance, transparency, and ethical considerations, ultimately shaping the trajectory of this rapidly evolving field.