New York State Prohibits DeepSeek on Government Devices

February 10, 2025 TruthUSA

This follows similar prohibitions in Texas, Taiwan, Australia, South Korea, Canada, the Netherlands, and Italy.

The Chinese AI chatbot DeepSeek has been prohibited from use on New York State’s government devices and networks, as announced by Governor Kathy Hochul on Monday.

The ban was implemented due to “serious concerns” about the app’s “link to foreign government surveillance and censorship, including its capabilities to harvest user data and compromise technological secrets,” according to a statement from Hochul’s office.

“The safety of the public is my highest priority, and we are taking decisive actions to shield New Yorkers from both foreign and domestic threats,” Hochul commented. “New York will persevere in combating cyber threats, protecting the privacy and security of our data, and defending against state-sponsored censorship.”

DeepSeek, an AI startup founded in Zhejiang, a province in southern China, alarmed AI investors last month after rolling out DeepSeek-R1 in January, claiming the chatbot matched the performance of OpenAI’s ChatGPT-01 at a fraction of the cost.

However, DeepSeek’s extensive data collection practices and its alignment with the censorship ideologies of the Chinese Communist Party (CCP) have spurred significant concerns, resulting in a series of bans regarding the app on government devices in Texas, Taiwan, Australia, South Korea, Canada, and the Netherlands.

Staffers in the U.S. House and Navy personnel have been cautioned against utilizing the app.

On February 7, House lawmakers proposed legislation to impose a federal ban on DeepSeek within executive agencies.

Italy has gone further than merely banning the app from government devices; the country’s data regulator initiated an investigation into the chatbot service on January 30. Consequently, the app remains unavailable on Google Play and Apple’s App Store in Italy.

According to Chinese intelligence law, all individuals and organizations must provide data to the state upon request for national security necessities.

This law has raised alarms in the United States and other nations regarding data controlled by services associated with China, including TikTok, a popular video-sharing platform.

On Sunday, South Korea’s intelligence agency stated that DeepSeek is “excessively” gathering personal data and utilizing all input data for its training processes.

“In contrast to other generative AI services, it has been confirmed that chat records are transferable since it has a function to collect keyboard input patterns that can identify individuals and communicate with servers owned by Chinese firms,” the National Intelligence Service of Seoul remarked.

The agency highlighted that DeepSeek grants advertisers unrestricted access to user data and retains South Korean users’ data on Chinese servers.

A Canadian cybersecurity firm, Feroot Security, reported that researchers uncovered code on DeepSeek’s login page capable of transmitting user data to China Mobile, a leading state-owned Chinese telecom enterprise prohibited from operating in the United States.

At the same time, analyses conducted by The Epoch Times have suggested that the application censors answers to inquiries about protests and human rights violations in China, further promoting the narratives of the CCP regarding Taiwan.

DeepSeek has not yet responded to a request for comment from The Epoch Times.

In response to the South Korean government’s ban on DeepSeek in select government sectors, a spokesperson for China’s Foreign Ministry, Guo Jiakun, stated on February 6 that Beijing “values data privacy and security highly and protects it in accordance with the law,” adding that the Chinese government “has never asked and will never ask any company or individual to collect or store data in violation of the law.”

Also on Monday, Demis Hassabis, co-founder and CEO of Google DeepMind, expressed to Bloomberg TV that he considers DeepSeek’s claims, particularly regarding the cost of its AI model, to be “exaggerated and somewhat misleading.”

He noted that DeepSeek seemingly reported “merely their final training run, which is simply a fraction of the overall expense typically required to explore, train, and conduct all the preliminary tests,” adding that he found no new technology within the Chinese model.

Source link