Microsoft restricts its employees from using DeepSeek AI

Microsoft has reportedly banned its employees from using the DeepSeek AI chatbot due to concerns over data security and potential propaganda influence.

Speaking at a Senate hearing on winning the AI race with China, Microsoft Vice Chairman and President Brad Smith disclosed that the tech giant doesn't allow employees to use the China-based AI app.

This is mainly because the user data is stored on Chinese servers, subject to Chinese law and intelligence agencies.

Additionally, DeepSeek censors sensitive topics as per Chinese government requirements.

Notably, this restriction is not surprising given DeepSeek's privacy policy. That said, it's notable that Microsoft had previously offered DeepSeek's R1 model on its Azure cloud service after conducting "rigorous red teaming and safety evaluations".

Smith also mentioned that Microsoft was able to modify DeepSeek's AI model to remove harmful side effects, but details of these modifications weren't disclosed.

For the unversed, China’s DeepSeek is open-source, allowing organisations to download and host the model themselves, potentially mitigating some risks.

On the other hand, Microsoft's own Copilot internet search chat app competes directly with DeepSeek's app, yet Microsoft's restriction seems driven by security concerns rather than competition, as other chat apps like Perplexity are available in the Windows app store.