Experts warn of privacy and security risks with Chinese AI model DeepSeek

Experts are raising concerns about privacy and potential misuse of user data as the Chinese AI model DeepSeek gains global popularity. Launched by a Hangzhou-based startup, DeepSeek has become the most downloaded free app on Apple’s App Store in the United States, rivalling ChatGPT. However, leading AI experts warn users to exercise caution when using the chatbot, particularly when sharing sensitive information.

Michael Wooldridge, a professor of AI at the University of Oxford, advised against inputting sensitive data into the chatbot. “I think it’s fine to download it and ask it about the performance of Liverpool football club or chat about the history of the Roman empire,” he told The Guardian. “But would I recommend putting anything sensitive, personal, or private on them? Absolutely not … Because you don’t know where the data goes.”

Cited in The Guardian report, Dame Wendy Hall, a United Nations AI advisor, pointed out the regulatory environment that Chinese tech companies operate in. “You can’t get away from the fact that if you are a Chinese tech company dealing with information, you are subject to the Chinese government’s rules on what you can and cannot say,” she said. Concerns have arisen that data input into DeepSeek’s systems could be accessible to the Chinese government due to laws requiring companies to cooperate with state authorities.

Ross Burley, co-founder of the Centre for Information Resilience, expressed alarm over how technologies like DeepSeek could be used for surveillance or influence campaigns. “We’ve seen time and again how Beijing weaponises its tech dominance for surveillance, control and coercion, both domestically and abroad,” Burley said. He warned that if left unchecked, platforms like DeepSeek could “feed disinformation campaigns, erode public trust, and entrench authoritarian narratives within our democracies.”

The rise of DeepSeek has also drawn reactions from government officials. Peter Kyle, the UK technology secretary, urged caution, acknowledging that the app has censorship built into it. “I think people need to make their own choices about this right now, because we haven’t had time to fully understand it,” he said on The News Agents podcast. “This is a Chinese model that … has censorship built into it. So, it doesn’t have the kind of freedoms you would expect from other models at the moment.”

DeepSeek’s rise to prominence highlights tensions between technological innovation and data privacy. Experts and officials are urging users to remain vigilant when interacting with the platform and to think twice before sharing personal information. As Wooldridge put it: “Curiosity is fine, but caution is essential.”

While there has yet to be formal advice from Australia's national security agencies about its use, Housing Minister Clare O'Neil has urged caution. “The app only came out in the last couple of days … so what our national security agencies will be doing at the moment is having a look at the settings of the app and understanding more about how it works before it issues some formal guidance to Australians about care that they need to take,” she told Sunrise. “Look, the advice is always the same. You should always exercise caution. The thing I'd be thinking about really for Australians at home is your personal information. That is the most valuable asset that we have in some respects. Just be careful about what you're sharing.”

O'Neil, a former home affairs minister, then mentioned she is a “ChatGPT girl” herself, raising questions about government ministers' use of AI. “Like many Australians, AI is going to be a tool that helps us do all the jobs that we do every day. From the most mundane things like typing in what food you have got in the fridge and asking it what recipes you can cook your kids for dinner. There is, you know, everything we do can be made simpler, easier and faster through AI.”