Microsoft has announced that its former partner, DeepSeek, will now have access to its reasoning model, R1, on Azure AI Foundry. The move follows reports of potential IP theft or service violations between the companies. R1, which was previously blocked by Chinese AI models due to censorship, is now available for use under strict evaluations including automated behavior assessments and security reviews.
Customers can soon use a simplified version of R1 locally on Copilot+ PCs that meet specific AI readiness criteria. While Microsoft emphasizes the model’s rigorously tested nature, some industry observers express concern over its accuracy and potential bias towards Chinese topics, as shown in recent tests by NewsGuard.
TechCrunch readers are encouraged to stay updated on the latest developments in AI technology.
**Title:** Microsoft Brings DeepSeek’s AI Model to Azure Despite Concerns
**
Source: https://techcrunch.com/2025/01/29/microsoft-brings-a-deepseek-model-to-its-cloud