Both OpenAI and Microsoft are reportedly investigating the Chinese startup to determine whether it accessed and utilized OpenAI's technology to develop its R1 reasoning large language model.

Figure 1. Microsoft Doubts DeepSeek at First, Then Integrates It into Its Cloud.

Despite launching an investigation into the Chinese AI startup DeepSeek, Microsoft has included the startup's latest reasoning model, R1, in its model catalog on Azure AI Foundry and GitHub. Figure 1 shows Microsoft Doubts DeepSeek at First, Then Integrates It into Its Cloud.

Azure AI Foundry, the rebranded version of Azure AI Studio, consists of the Azure AI Foundry portal (formerly known as Azure AI Studio), the Azure AI Foundry software development kit (SDK), Azure AI Agents, pre-built app templates, and various tools for AI application development.

In a blog post announcing the addition of R1, Microsoft stated that the model had “undergone rigorous red teaming and safety evaluations, including automated assessments of model behavior and extensive security reviews to mitigate potential risks.”

The Microsoft blog highlighted that customers will soon be able to use distilled versions of the R1 model to run locally on their Copilot+ PCs.

By accessing the R1 model via Azure AI Foundry and GitHub, enterprises can experiment with, compare, and implement the model in their code for free, either in the playground or via the API.

However, despite adding the R1 model to its cloud platform, Microsoft and OpenAI are reportedly investigating the startup behind the R1 model for potential intellectual property (IP) infringement.

A separate Bloomberg report stated that both US-based companies are probing whether DeepSeek accessed and copied OpenAI’s technology while developing its reasoning model. The report further mentions that Microsoft’s security researchers observed individuals associated with DeepSeek extracting large amounts of data using OpenAI’s API, the tool designed for interacting with OpenAI services or models.

As an investor in OpenAI, Microsoft reportedly informed the company about the alleged suspicious activity.

Another reason for the investigation is the growing popularity of DeepSeek's model among users and cloud providers. Upon its release, the DeepSeek model quickly surpassed ChatGPT—OpenAI’s assistant—in terms of downloads and ranking on the Apple App Store.

The White House also seems to suspect that DeepSeek may have found a way to copy OpenAI’s technology to create its R1 model. In an interview with Fox News, David Sacks, often referred to as the White House’s AI czar, suggested that DeepSeek may have infringed upon OpenAI's intellectual property. "There’s substantial evidence that what DeepSeek did here is they distilled the knowledge out of OpenAI’s models," Sacks was quoted as saying.

Furthermore, Reuters reported that an OpenAI spokesperson stated that Chinese companies are consistently attempting to replicate models from leading US-based AI providers.

In addition to the probe by Microsoft and OpenAI, both White House and European regulators are investigating DeepSeek over national security risks, data privacy concerns, and potential IP violations. The White House confirmed on Tuesday that the National Security Council (NSC) is reviewing the AI model’s implications amid fears that Chinese AI advancements could challenge the dominance of US-based AI companies like OpenAI and Google. Meanwhile, Italy's data protection authority, the Garante, has launched its own investigation into DeepSeek, demanding transparency regarding its data collection practices.

Source: NETWORK WORLD

Cite this article:

Priyadharshini S (2025),"Microsoft Initially Questions Deepseek, Then Integrates It into Its Cloud",Anatechmaz ,pp.122