Artificial intelligence models developed in China and provided to Western businesses and governments at low or no cost are raising concerns among experts. Chinese AI developers, including DeepSeek and Xiaomi, regularly release 'open-weight' models, which allow users to utilise the core software without paying ongoing licence fees.
However, Ross Burley, director of the Centre for Information Resilience (CIR), stated that running a model on a user's own servers does not automatically remove political censorship or repair security weaknesses. An evaluation report by the US National Institute of Standards and Technology (Nist) identified political censorship in downloaded DeepSeek models operating on its own infrastructure.
Michael Wooldridge, a professor of AI at the University of Oxford, suggested it is reasonable to assume that data entered into DeepSeek's chatbot could be shared with the Chinese state. DeepSeek's online chatbot policy indicates it collects prompts and uploaded material, processes and stores personal data in China, and permits disclosure to Chinese authorities.