Despite ongoing disputes with OpenAI concerning alleged IP theft, Microsoft has released DeepSeek’s reasoning model, R1, on its Azure AI Foundry platform. This marks a significant step for Microsoft’s integrated AI services for organisations, following intensive evaluations to ensure security and functionality.
In the imminent future, pared-down versions of R1 will be available to run locally on Microsoft’s Copilot+ PCs that meet specific AI readiness criteria. Microsoft anticipates this will enable developers and corporations to tackle complex challenges and offer transformative experiences.
The deployment of R1 on Microsoft’s cloud service, Azure, raises questions, given the recent scrutiny of DeepSeek for possibly misusing OpenAI and Microsoft’s services. While there are still uncertainties circulating around DeepSeek’s actions, the integration of the popular R1 model to the cloud services may prove beneficial.
Lastly, it’s worth noting that Microsoft’s potential enhancements to the R1 model, and its possible censorship, still remain ambiguous. Reports suggest R1 often provides inadequate responses or avoids answering altogether, specifically focused on news-related topics and China associated queries, likely due to China’s strict AI development regulations.
Original source: Read the full article on TechCrunch
