TAIPEI, Taiwan – Taiwan said its public sector workers would be banned from using DeepSeek, a recently released Chinese artificial intelligence model, due to concerns over data leaks, joining a group of U.S. government agencies that reportedly introduced a similar restriction.
DeepSeek is a Chinese artificial intelligence company, founded in 2023, that develops open-source large language models. Its AI chatbot app has gained significant popularity, surpassing ChatGPT as the most-downloaded free app on the iOS App Store in the United States by Jan. 27.
Despite its success, concerns have been raised regarding DeepSeek’s ties to the Chinese government, data security, and potential censorship within its AI models.
“Government agencies and critical infrastructure should not use DeepSeek, because it endangers national information security,” said Taiwan’s Ministry of Digital Affairs in a Friday statement.
“DeepSeek AI service is a Chinese product. Its operation involves cross-border transmission, and information leakage and other information security concerns,” it added.
The ministry said the ban applied to employees of central and local government agencies, public schools, state-owned enterprises and other semi-official organizations, as well as those working on “critical infrastructure projects” and in government-endowed foundations, but did not elaborate on specifics or enforcement measures.
Wang Ting-yu, a legislator from the island’s ruling Democratic Progressive Party, also called on businesses and individuals in the private sector to be “more vigilant” about cybersecurity and avoid using DeepSeek products if there are security risks.
Taiwan’s announcement follows a similar restriction by some U.S. government agencies, including the Navy.
The Navy instructed its members to avoid using DeepSeek, saying the service was not to be used “in any capacity” due to “potential security and ethical concerns associated with the model’s origin and usage,” according to media reports.
NASA also reportedly banned the use of DeepSeek on Friday, informing its employees servers of the Chinese AI chatbot “operate outside of the United States, raising national security and privacy concerns.”
Earlier, media reported that U.S. congressional offices were being told that use of DeepSeek was “unauthorized for official House use.”
Japan, South Korea
Masaaki Taira, Japan’s digital transformation minister, said on Friday that public officials should refrain from using DeepSeek to protect data.
“The question is whether personal information is properly protected … Our Personal Information Protection Commission should respond if necessary from the perspective of protecting users’ private information,” Taira said during a speech in Okinawa.
Taira also urged all the government organizations to be alerted, hinting at a policy to enforce such a measure, without elaborating.
In January, Itsunori Onodera, a legislator from Japan’s ruling Liberal Democratic Party, urged the public to stop downloading DeekSeep, calling it “dangerous,” citing the AI chatbot’s response regarding a territorial dispute between China and Japan over islands known in China as the Diaoyu Islands and in Japan as the Senkaku Islands.
DeepSeek said the disputed islands had “always been an inherent part of Chinese territory, over which China has indisputable sovereignty.”
In response, Japanese Prime Minister Shigeru Ishiba said: “We need to develop and use reliable AI.”
RELATED STORIES
How does DeepSeek answer sensitive questions about China?
China’s DeepSeek has close ties to Beijing
China’s DeepSeek AI tops app charts in US, Europe
In South Korea, the Personal Information Protection Commission said it would send an official inquiry to the DeepSeek headquarters in China to confirm its procedures for collecting personal information, and how it is processed and stored.
The commission added it would also question how collected personal information was used and details about the chatbot’s AI learning process.
“This is a measure to address the growing concerns over the rapid increase in the number of domestic users of DeepSeek. We will review the proposed response and, if necessary, begin an inspection or investigation,” the commission said.
Europe, UK
The Irish Data Protection Commission said it had sent a letter to DeepSeek requesting information about the processing of data relating to Irish users, amid concerns about possible violations of the European Union’s personal information-related laws.
The British government also said it was looking at DeepSeek from a national security perspective, with media reporting that Italy blocked access to DeepSeek’s application over data privacy concerns and initiated an investigation into the company’s data handling practices, and that German authorities were looking into the possibility of taking regulatory action against the AI chatbot.
DeepSeek has not issued a public response to the global security concerns raised about its AI applications.
Edited by Mike Firn.