South Korean ministries, police block DeepSeek access
1. Seoul, along with countries such as France, Italy, have asked questions about Chinese AI startup's data practices
South Korea Blocks DeepSeek Access Over Data Privacy Concerns
SEOUL: South Korean ministries and law enforcement agencies announced Thursday that they are restricting access to DeepSeek on government work computers after the Chinese AI startup failed to respond to inquiries from the country’s data watchdog regarding its handling of user information.
DeepSeek, which claims to rival top U.S. artificial intelligence firms at a fraction of the cost, launched its R1 chatbot last month, positioning itself as a disruptor in the global AI industry. However, concerns over its data management practices have drawn scrutiny from multiple governments, including South Korea, France, and Italy.
South Korea’s Personal Information Protection Commission had submitted a formal request for details on DeepSeek’s data handling policies. But after the company did not provide a response, various ministries confirmed they were taking steps to limit access to prevent potential security risks posed by generative AI technologies.
A spokesperson for South Korea’s defense ministry told AFP that it had implemented blocking measures specifically for military work-related PCs connected to the internet. "We have also reiterated security precautions to all units and personnel regarding the use of generative AI, considering both security and technical risks," the official added.
The National Police Agency confirmed that it had also blocked access to DeepSeek, while the trade ministry said it had temporarily restricted usage across all its computers, citing the company’s failure to respond to the watchdog’s inquiry. The finance ministry similarly stated that it had introduced measures to prevent the leakage of personal and confidential information through DeepSeek for all employees.
International Crackdown on DeepSeek
South Korea’s actions come amid growing concerns worldwide about DeepSeek’s data practices. Last week, Italy launched an investigation into the company’s R1 model and blocked it from processing Italian users’ data. Australia has also banned DeepSeek on all government devices based on security agency recommendations.
Kim Jong-hwa, a professor at Cheju Halla University’s artificial intelligence department, suggested that geopolitical tensions between the U.S. and China might be influencing responses to DeepSeek. However, he maintained that restrictions were warranted.
"From a technical standpoint, AI models like ChatGPT also face unresolved security challenges," Kim told AFP. "Given that China operates under a communist regime, I question whether they prioritize security concerns as much as OpenAI does when developing AI technologies."
"We currently lack sufficient insight into how much emphasis DeepSeek has placed on security when building its chatbot. Therefore, taking precautionary measures is not excessive," he added.
DeepSeek has stated that it developed its AI model using H800 chips, which were legally available for sale to China until 2023 under U.S. export controls.
South Korea Boosts AI Investment Amid Rising Competition
The DeepSeek controversy unfolds as South Korea ramps up investments in its semiconductor and AI industries. On Wednesday, the government announced an additional 34 trillion won ($23.5 billion) for high-tech sectors, with acting President Choi Sang-mok urging domestic companies to remain competitive.
"Recently, a Chinese company introduced DeepSeek R1, a high-performance AI model with low costs, creating fresh waves in the market," Choi said. "The global AI race is shifting beyond infrastructure scaling to a more complex competition involving software capabilities and other factors."
South Korean tech giants Samsung Electronics and SK hynix, key suppliers of advanced AI chips, are expected to play a critical role in shaping this evolving landscape.
