DeepSeek has been banned on government devices in South Korea, Australia, and Taiwan—with additional countries possibly following suit soon.
In recent days, various governmental bodies, including those in South Korea and Australia, have blocked employees from accessing Chinese AI startup DeepSeek’s chatbot application, largely due to data security issues.
Countries such as the United States are reportedly considering similar restrictions, citing concerns over the Chinese tech’s data practices and lack of transparency around user privacy.
DeepSeek garnered major attention recently after making claims about its low-cost model, which allegedly delivers competitive performance at a fraction of the investment of Western AI models—causing U.S. tech stocks to fall sharply. Social media humorously commented: “ChatGPT just got replaced by AI.”
Here’s a deeper look at DeepSeek and the reasons behind the global pushback.
What is DeepSeek?
DeepSeek is a Chinese artificial intelligence company headquartered in Hangzhou. It was established in 2023 by Liang Feng, who previously co-founded the hedge fund High-Flyer, valued at $7 billion, with university friends back in 2016.
In January, DeepSeek released its latest AI chatbot model, DeepSeek R1, which is free to use and resembles OpenAI’s ChatGPT in design and functionality.
These AI chatbots simulate natural human conversation, allowing users to input queries and receive AI-generated responses. The bots can be used for various tasks, such as solving equations, drafting emails or documents, translation, and coding support.
How is DeepSeek different from ChatGPT and similar platforms?
A key distinction is cost-efficiency. According to DeepSeek, their latest model required less than $6 million in Nvidia H800 chip power to train.
That’s a fraction of the billions spent by U.S. tech giants—OpenAI reportedly invested over $100 million to develop GPT-4, according to Wired in 2023.
DeepSeek’s app launched publicly on January 20 and soared to the top of Apple’s App Store within a week. This sudden rise spooked investors, with Nvidia losing nearly $600 billion in market capitalization after a 17% share drop.
Which countries have moved to ban DeepSeek?
A number of countries have taken steps to ban DeepSeek from government systems and devices:
United States
The Wall Street Journal reported that U.S. lawmakers intend to introduce legislation blocking DeepSeek on federal devices.
NASA banned the tool on January 31, and the U.S. Navy had already advised staff not to use it due to potential security and ethical concerns stemming from its Chinese origin.
South Korea
South Korea’s Ministry of Trade, Industry and Energy temporarily prohibited DeepSeek on government devices, citing security concerns.
The nation’s Personal Information Protection Commission also requested detailed information from DeepSeek on January 31 regarding user data handling.
Korea Hydro & Nuclear Power also blocked AI apps, including DeepSeek, on employee devices.
Australia
The Australian Department of Home Affairs issued a directive banning DeepSeek across all government systems and devices due to national security concerns.
Home Affairs Minister Tony Burke confirmed the move was to protect national interests and sovereignty.
Italy
On January 30, Italy’s data regulator, Garante, restricted DeepSeek from processing Italian user data, citing lack of transparency.
Earlier, the authority requested clarity on how DeepSeek handles and stores user information.
Taiwan
On Monday, Taiwan blocked DeepSeek for government use, also referencing security risks tied to the app’s origin.
Why are countries banning DeepSeek?
Most governments cite national security risks tied to the Chinese-developed app and an absence of clear information about how it manages personal data.
DeepSeek’s privacy policy states it collects:
- Personal information like email, phone number, password, and date of birth;
- Chat history, including voice and text inputs;
- Device and network info such as IP address and typing behavior;
- Data may be shared with advertisers and service providers, retained indefinitely.
OpenAI’s ChatGPT policy is similar, collecting registration and device data, and storing chatbot inputs “as long as needed.” However, OpenAI clarifies that this data may be shared only with affiliates.
A report cited by ABC News on Wednesday included claims from Feroot Security CEO Ivan Tsarynny that DeepSeek’s code includes a mechanism to transmit user data to CMPassport.com, a Chinese government-affiliated registry.
This claim has yet to be independently verified by Al Jazeera.
Is this level of data collection unique?
Eddy Borges-Rey of Northwestern University in Qatar explains that data harvesting is widespread among big tech firms—Meta, Google, and OpenAI all gather user data for ads, algorithm training, and AI improvements.
While many Western firms have faced fines or probes for data misuse, they remain operational under regulatory frameworks like the EU’s GDPR.
Borges-Rey points out that Western governments view Chinese tech firms differently due to Chinese laws such as the National Intelligence Law, which allows state access to corporate data. This raises alarms over potential misuse for espionage or influence operations.
While it’s unclear whether such practices are actually taking place, the possibility alone justifies preemptive government bans in the West.
Western platforms are often regulated but not viewed as national security threats by their own governments, unlike Chinese counterparts.
Are there other concerns around AI chatbots?
Yes, aside from national security and data privacy, there are growing concerns in education and workplace environments, says Nayana Prakash from Chatham House.
These include:
- Biases in AI responses;
- Copyright violations;
- Threats to original content and creative work due to the proliferation of large language models (LLMs).
China, for its part, bans a range of Western apps—including Facebook, X, and ChatGPT—for all users within its borders.
