As governments grapple with what to do about DeepSeek, China’s hot new AI model — including outright bans — the UK is issuing a warning to its citizens and business: be careful.

“Ultimately it is a personal choice for people, whether they decide to download it or not,” UK AI minister Feryal Clark said in an interview with Bloomberg on Friday. “My advice will be to make sure if people are downloading it that they are alert to the potential risks and they know how their data will be used.”

DeepSeek sent shockwaves through global markets this week after the Chinese-made AI model showed results comparable to the top US-made systems, but developed at what the company describes as a fraction of the cost. The model’s mobile app proved popular, topping download rankings. It has prompted various western countries to consider the national security implications of a cheap and popular Chinese-made model.

DeepSeek’s AI Restricted by ‘Hundreds’ of Companies in Days

The UK’s approach has been less restrictive. Government officials have been privately asking businesses to be careful about the information they share with DeepSeek and other Chinese AI apps over concerns that it could be accessed by the Chinese state, according to people familiar with the matter, who asked not to be identified while discussing conversations that haven’t been made public.

The UK’s warnings are more passive than the restrictive measures others have taken. Italy’s privacy regulator blocked the app, while the US Pentagon has also restricted access. Hundreds of companies, including law firms, have also stopped their staff from accessing the DeepSeek’s tools.

The rise of DeepSeek runs parallel to the concerns around TikTok’s connections to China. UK officials banned the app on government phones and fined the company for safety breaches, but stopped shorted of following the US with a complete ban. Prime Minister Keir Starmer’s Labour government is looking to reset relations with China as part of its search for economic growth.

Clark said she won’t be downloading the app herself, but the UK will monitor developments around the AI model.

The UK has established an institute to help minimize shocks from rapid and unexpected advances in the technology. This week it will begin the process of creating what it hopes will be a new global standard for AI cyber security.

The country will submit a new voluntary code of practice for businesses and public services to follow to the European Telecommunications Standards Institute. Government officials say following this code will help protect developers against cyber-threats, hacking and system failures.

“As we look to diffuse the technology across government and private sector, we want to make sure the companies who are developing the technology are thinking about cybersecurity from the start,” Clark said.

The new code has been long planned by the UK and is unrelated to the emergence of DeepSeek. The UK is also working on a cybersecurity bill, announced last year, to address vulnerabilities in the UK’s networks. Clark warned that threats from both malicious actors and nation states have increased in recent years.

Photograph: The UK establishes an institute to help minimize shocks from rapid and unexpected advances in the technology. Photo credit: Lam Yik/Bloomberg

Related:

Copyright 2025 Bloomberg.