Microsoft has added the R1 model from the Chinese company DeepSeek to its Azure AI Foundry platform. This model has already generated significant interest in US financial circles, as its training costs are lower than those of leading models from OpenAI. R1 is now part of the model catalog on Azure AI Foundry and GitHub, allowing Microsoft customers to integrate it into their applications.
Microsoft announced that the R1 model has undergone thorough safety checks and behavioral assessments. In the future, users will be able to use a simplified version of R1 on Copilot Plus computers. This will enable developers to experiment and integrate AI into their workflows more quickly.
However, Microsoft and OpenAI are investigating a possible violation of terms of use by DeepSeek. The investigation began after Microsoft security researchers detected significant data usage via the OpenAI API, which may have been linked to DeepSeek.
Despite these issues, Microsoft has integrated R1 into its cloud services, as the model continues to attract user attention. However, there are doubts about R1’s accuracy, as testing has shown it often provides inaccurate answers or refuses to respond to queries related to certain topics.
