Microsoft Adds DeepSeek’s R1 AI Model to Azure Despite OpenAI Probe

Despite ongoing concerns about DeepSeek’s potential misuse of OpenAI’s technology, Microsoft has added DeepSeek’s R1 AI model to its Azure AI Foundry service.

Azure AI Foundry is Microsoft’s platform for enterprise AI models, providing developers with access to a range of AI solutions. In a blog post, Microsoft emphasized that R1 has undergone rigorous security testing, including red teaming and automated assessments of model behavior to mitigate risks.

DeepSeek Under Scrutiny

The decision to integrate R1 into Microsoft’s cloud comes amid reports that Microsoft and OpenAI are investigating DeepSeek for potential IP violations. According to Bloomberg, Microsoft’s security team detected unusual activity in late 2024, suggesting that DeepSeek may have extracted a significant amount of data using OpenAI’s API. Microsoft, OpenAI’s largest investor, alerted OpenAI to the possible breach. You can read more about this here.

Despite these concerns, R1 has gained significant attention for its reasoning capabilities, making it an attractive addition to Microsoft’s growing AI ecosystem.

Expanding Access to R1

Microsoft also revealed that in the near future, "distilled" versions of R1 will be available on Copilot+ PCs—its AI-ready Windows hardware. This move aligns with Microsoft’s broader strategy of bringing more powerful AI tools directly to users.

"As we continue expanding the model catalog in Azure AI Foundry, we’re excited to see how developers and enterprises leverage […] R1 to tackle real-world challenges and deliver transformative experiences," Microsoft stated in its blog post.

Lingering Questions About R1’s Reliability

While R1 is making headlines, questions remain about its accuracy and censorship.

• A NewsGuard study found that R1 provided inaccurate answers or non-answers 83% of the time when asked about news-related topics.

• Another study revealed that R1 refuses to respond to 85% of prompts related to China, likely due to government censorship restrictions on AI models developed in the country.

It’s unclear whether Microsoft has modified the model to improve its accuracy or address these censorship issues.

Looking Ahead: A Strategic but Risky Move

Microsoft’s decision to host R1 on Azure signals a willingness to embrace new AI models—even amid controversy. While DeepSeek’s model is undeniably powerful, the ongoing investigation into its data practices could impact its long-term viability.

For now, Microsoft is betting that R1’s capabilities will outweigh any potential risks. Whether that gamble pays off remains to be seen.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.