DeepSeek usage a ‘personal choice’ for public, Government says
17 February 2025, 18:14
It comes after South Korea became the latest nation to limit access to the China-based AI chatbot.
Downloading Chinese AI chatbot DeepSeek is a “personal choice” for the public, the Government has said, adding that it keeps new technologies under constant review.
South Korea has become the latest country to limit access to the new chatbot, pausing new downloads of DeepSeek from app stores in the country over privacy concerns.
DeepSeek shook the AI market when it emerged earlier this year, claiming to offer similar performance to OpenAI’s market-leading ChatGPT for a fraction of the development cost, but the company’s ties to China – including storing data on servers in the country – have raised privacy and security concerns around the world.
There are fears that personal data gathered from inputs into DeepSeek and stored on servers in China could be accessed by the Beijing government for surveillance or intelligence.
South Korea’s intervention came after Taiwan and Australia banned DeepSeek from government devices, while Italy has banned the app outright and legislators at federal and state level in the US have proposed legislation for a ban over surveillance concerns.
Asked about the UK’s stance, a spokesperson from the Department for Science, Innovation and Technology said: “The use of DeepSeek is a personal choice for members of the public.
“AI technologies have the power to transform our society, but any new tools that become available must have security built in from the outset to keep user data safe.
“Through our new blueprint for AI, we have also set out a bold vision to ensure that Britain becomes a global magnet for innovation – ensuring future waves of game-changing technology are developed on our shores.”
The Government has strict rules around apps allowed on official devices. TikTok, which also has links to China, is banned on such devices in the UK, and new apps are technologies are kept under review as they continue to evolve.
Bill Conner, chief executive of AI software firm Jitterbit, and a former security adviser to the US and UK governments, said the public and businesses around the world should seriously consider whether to use the app before downloading it, or accessing the chatbot through the website.
“DeepSeek potentially presents a new level of threat to enterprises, businesses and governments globally,” he said.
“Its seemingly overnight popularity and free-to-use AI model make it look like an innovative new arrival and serious swap-out competitor to well-funded LLMs (large language models) from its well-trusted rivals.
“In reality, DeepSeek represents a clear risk for any enterprise whose leadership values data privacy, security and transparency.
“Any time you have AI models with data servers built in China, by nature there are privacy concerns due to the fact of where the data is housed and what biases are built into the AI models.
“I would recommend business and technology leaders conduct their own due diligence to assess exactly where their data is stored, how it is being used as well as any AI modelling biases.
“Each company should assess their own compliance, governance and security risks associated with running AI models and data.”
