Microsoft has officially prohibited its employees from using the China-based AI application DeepSeek, citing serious concerns about data privacy and potential propaganda content.
During a recent hearing in the U.S. Senate, Microsoft Vice Chair and President Brad Smith announced the decision, highlighting the risks associated with how DeepSeek handles user data. According to Smith, user information collected by the app is stored on servers located in China — a country where local laws require such data to be shared with intelligence agencies upon request.
Another key concern raised is DeepSeek’s content filtering mechanisms. The application reportedly censors information on topics the Chinese government deems “sensitive,” raising alarms about its objectivity and potential to spread state-influenced narratives.
Earlier this year, DeepSeek’s R1 model gained significant attention, leading Microsoft to integrate it into its Azure cloud platform. Thanks to the model’s open-source nature, Microsoft was able to run it on its own servers, preventing data from being transmitted to China. However, this safeguard doesn’t completely eliminate the risk of the model generating biased or potentially harmful content.
Smith emphasized that before deploying DeepSeek’s model on Azure, Microsoft conducted thorough evaluations to remove any “harmful side effects.” Still, the company’s strict stance reflects a growing global caution toward AI tools developed in China — particularly when data security and information integrity are at stake.
Leave a Reply