DeepSeek's R1 Now Accessible on Azure AI Foundry and GitHub

Microsoft is continuing to work with DeepSeek, even though OpenAI, a close partner, might be suggesting that DeepSeek used its intellectual property without permission. Despite this ongoing investigation, Microsoft believes in the value of DeepSeek’s AI technology and is moving forward with adding its models to its cloud platform.

R1 AI Model Now on Azure AI Foundry

Microsoft has made DeepSeek's R1 reasoning AI model available on its Azure AI Foundry platform. This platform brings different AI tools for businesses into one place, and R1 has gone through safety checks and security reviews to make sure it meets Microsoft’s standards.

R1 to Work on Microsoft’s Copilot+ PCs

Soon, Microsoft plans to offer simpler versions of R1 that can be run directly on its Copilot+ PCs, which are specially designed computers ready for AI tasks. This will let customers use AI models on their own devices, offering more flexibility.

Concerns About Data Misuse and R1’s Accuracy

While Microsoft is adding R1 to its services, it is also looking into whether DeepSeek misused OpenAI’s data. Tests of the R1 model show that it often gives wrong or incomplete answers, especially about news, and avoids answering questions related to China, possibly due to government censorship. Microsoft is addressing these concerns as it moves forward with R1.

The official link to this news is https://azure.microsoft.com/en-us/blog/deepseek-r1-is-now-available-on-azure-ai-foundry-and-github/.