Microsoft Bans DeepSeek App for Employees
Microsoft has banned its employees from using the DeepSeek AI app. This announcement came from Microsoft Vice Chairman and President Brad Smith during a recent Senate hearing on AI.
Smith cited data security and the potential for Chinese propaganda influence as key reasons for the ban. He confirmed that Microsoft also hasn't included DeepSeek in its app store due to these concerns.
Data Security and Chinese Influence
The ban stems from DeepSeek's storage of user data on Chinese servers. This data is subject to Chinese law, requiring cooperation with intelligence agencies. DeepSeek also censors topics deemed sensitive by the Chinese government.
At Microsoft, we don't allow our employees to use the DeepSeek app.
This restriction marks the first public acknowledgment of a DeepSeek ban by Microsoft, although other organizations and countries have implemented similar restrictions.
Microsoft's Conflicting Actions
Despite the ban, Microsoft previously offered DeepSeek's R1 model on its Azure cloud service. This offering differs from the banned app, as the open-source nature of DeepSeek allows anyone to host the model independently, bypassing data transfer to China.
However, this doesn't eliminate risks associated with potential propaganda or the generation of insecure code. Smith stated that Microsoft modified DeepSeek's model on Azure to mitigate "harmful side effects," but provided no further details.
Microsoft claims to have conducted rigorous red teaming and safety evaluations before offering DeepSeek R1 on Azure.
Competition and App Store Availability
DeepSeek's app directly competes with Microsoft's Copilot. While Microsoft hasn't banned all competing chat apps from its Windows app store, apps from Google, including its chatbot Gemini, were not found during a recent search.
The ban highlights the complex relationship between US tech companies and Chinese AI developments, raising important questions about data security, censorship, and international competition.
