Exciting news has emerged as Microsoft recently announced that R1, the reasoning model known as DeepSeek, will soon be accessible on the Azure AI Foundry platform. This service consolidates various AI offerings for businesses under one unified brand.
According to their official blog post, Microsoft revealed that the R1 version on Azure AI Foundry has undergone “rigorous security evaluations and team collaborations,” including “automated model behavior assessments and extensive safety reviews to mitigate potential risks.”
In the near future, users will be able to leverage a “distilled” version of R1 to run locally on Copilot+ PCs.
Read Also:
However, the inclusion of R1 in Microsoft’s cloud services raises eyebrows, especially since Microsoft was reportedly investigating potential misuse of its services and OpenAI’s API by DeepSeek. Security researchers working with Microsoft suggested that DeepSeek might have stolen significant data using OpenAI’s API in the fall of 2024. As the largest shareholder of OpenAI, Microsoft informed them about these suspicious activities, as reported by Bloomberg and TechCrunch.
Still, given R1’s growing popularity, Microsoft may be keen to integrate it into its cloud offerings while DeepSeek remains in the spotlight.
For now, it remains unclear whether Microsoft has made any modifications to the model to improve its accuracy.
According to tests conducted by the reliable NewsGuard organization, R1 provided inaccurate or nonresponsive answers 83% of the time when queried about news-related topics. Additionally, a separate study found that R1 refused to answer 85% of questions related to China, likely due to the government censorship impacting AI models developed in the country.
Details about this announcement can be found on Azure’s official page, where Azure customers can now access DeepSeek in the available model catalog.
What are your thoughts? Share your opinions in the comments section below.
Via: TechCrunch
