The ban was made to protect ‘national information security,’ Taiwanese Premier Cho Jung-tai said.
Taiwan has told public agencies not to use the artificial intelligence (AI) service from DeepSeek, joining a growing list of countries that are examining the China-controlled application.
Taiwanese Premier Cho Jung-tai announced the ban on DeepSeek during a cabinet meeting on Feb. 3, stating that the decision was made to “protect the national information security,” according to a
statement from his office.
An exception has been made for colleges and research institutions accessing DeepSeek’s AI for research purposes. Wu Cheng-wen, head of Taiwan’s top science and technology department, advised researchers to download the app only on devices that contain no personal data or research materials and to keep its use confined to offline environments.
There has been rising international concern over the Chinese Communist Party’s backing of DeepSeek, a fast-growing startup headquartered in Hangzhou but with most of its subsidiaries operating from Beijing. Major concerns center on potential security flaws,
censorship rules, and compliance with Chinese laws that could see users’ data end up in the hands of Beijing.
Last week, Taiwan’s Ministry of Digital Affairs issued a warning to government departments and critical infrastructure sectors, urging them against the use of DeepSeek’s AI service due to concerns over possible data leaks.
During the cabinet meeting, Cho acknowledged that the Taiwanese government is aware of the
copyright infringement concerns linked to the Chinese-made AI service, alongside skepticism about the data used by DeepSeek to train its language models, which raise potential issues of ideological censorship and bias.
Cabinet spokesperson Michelle Lee quoted the premier stating that the risks associated with sending data to China while using DeepSeek’s AI services ”have not been clarified, nor have they been completely ruled out at this time.”
Founded in 2023 by Chinese businessman Liang Wenfeng, DeepSeek
stated the company adheres to Chinese laws and regulations, as well as “socialist core values.” In its privacy policy, the company said the app collects data such as date of birth, email addresses, text and audio input, and other information, all of which are stored on servers located in China.
Chinese regulations and laws, especially the 2017 National Intelligence Law, compel individuals and companies to cooperate with Chinese authorities in their intelligence work, such as handing over data collected both inside and outside the country.
DeepSeek
jolted the global market earlier this month after it revealed its new language learning model, R1, claiming it rivals U.S. competitor OpenAI’s reasoning model, O1. The AI assistant application shot to the top of Apple’s App Store in the United States and UK on Jan. 27.
But on the same day, DeepSeek
said it would temporarily limit user registrations amid record-breaking interest in its app. It cited “large-scale malicious attacks” targeting its services as the reason, but users with Chinese phone numbers were still able to register.
DeepSeek’s rapid popularity has drawn scrutiny. Concerns over users’ privacy and security have prompted foreign regulators to take action.
Italy was the first country to block DeepSeek. On Jan. 30, Italy’s data protection authority announced it had mandated the Chinese companies behind the app, Hangzhou DeepSeek Artificial Intelligence and Beijing DeepSeek Artificial Intelligence, to stop processing Italian users’ data immediately. The regulator is currently investigating how DeepSeek collects and handles Italian users’ data.
In the United States,
House offices have been warned against using DeepSeek, according to an internal notice seen by The Epoch Times.
U.S. Navy service members were also advised not to use any open-source AI programs for official work, including the DeepSeek chatbot, a Navy spokesperson confirmed to The Epoch Times on Jan. 30
On Jan. 31, Texas Gov. Greg Abbott issued a state ban against DeepSeek along with five other China-based social media apps, restricting their use on devices issued by the state government.
“Texas will not allow the Chinese Communist Party to infiltrate our state’s critical infrastructure through data-harvesting AI and social media apps,” the Republican governor stated in a
statement. “Texas will continue to protect and defend our state from hostile foreign actors.”
Aldgra Fredly and Frank Fang contributed to this report.
