Experts urge caution over using DeepSeek AI chatbot because of China links
29 January 2025, 15:54
The new AI chatbot has emerged as a rival to US firms such as OpenAI.
Security experts are urging people to be cautious if considering using emerging AI chatbot DeepSeek because of the app’s links to China and the potential implications for personal data.
The chatbot has upended the AI market with its emergence as a rival to the likes of ChatGPT and OpenAI, despite having been apparently developed for a fraction of the budget of its big US rivals.
However, some experts have raised concerns about using the app because DeepSeek is based in China, and there could be “risks” to data privacy as a result.
A number of major Chinese firms and technologies have found themselves sanctioned in recent years – most notably Huawei and TikTok – because of fears of data being harvested by the Chinese government for intelligence purposes.
The US has already announced it is examining potential national security risks around DeepSeek.
Bill Conner, chief executive of automation firm Jitterbit, and a former security adviser to UK and US governments, said DeepSeek “represents a clear risk to any enterprise whose leadership values data privacy, security and transparency”.
“Proactive and privacy-minded enterprises should do strict due diligence with all large language models (LLMs) and AI services, not just DeepSeek,” he said.
“But in this case, and as stated in their own privacy policy, DeepSeek is a shared cloud service run in China with data being stored in China — potentially introducing unknown risks to data privacy, compliance mandates and security controls.”
Industry analysts at Forrester echoed this concerns, with several of its experts noting in a joint blog post that DeepSeek’s privacy policy acknowledged the app collected user data and could share it with “public authorities at its discretion”.
“The massive downloads of DeepSeek mean that thousands – and even millions of users – are experimenting and uploading what could be sensitive information into the app,” the analysts said.
“According to its privacy policy, DeepSeek explicitly says it can collect ‘your text or audio input, prompt, uploaded files, feedback, chat history, or other content’ and use it for training purposes.
“It also states it can share this information with law enforcement agencies and public authorities at its discretion.
“Educate and inform your employees on the ramifications of using this technology and inputting personal and company information into it.”
Professor Allen Tucker, professor of artificial intelligence and head of the intelligent data analysis group at Brunel University of London, said AI chatbots in general should be approached with caution.
“Both have been designed with no use-case in mind. They are essentially toys to make people go ‘wow’ rather than solve any genuine problem,” he said.
Comparing DeepSeek with market leader ChatGPT, Professor Tucker added: “They are also both deeply embedded in their countries’ politics and therefore need to be treated with a great deal of caution – try asking either of them about controversial topics and they will cover up, change topics or shut down the conversation.”
