Not Wanting To Lose, Microsoft Launches DeepSeeek AI Model In Azure

Illustration of Microsoft Azure (photo: dock. microsoft)

JAKARTA Microsoft officially launched the R1 artificial intelligence (AI) model from Chinese startup DeepSeek, on the Azure cloud computing platform as well as the GitHub developer tool. This move was announced by the US technology company on Wednesday 29 January.

The AI R1 model is now available in the Azure and GitHub model catalog, joining more than 1,800 other models that Microsoft has offered.

DeepSek has just launched a free AI assistant that is claimed to use less data at a lower cost than existing AI services. In just a matter of days, this app successfully surpassed the number of ChatGPT downloads from the Apple App Store, sparking panic among technology stock investors.

The launch of the DeepSeek AI model in Azure also shows Microsoft's efforts to reduce its dependence on OpenAI, the ChatGPT maker. Previously, Reuters reported that Microsoft was adding an internal AI model and a third party to strengthen its flagship AI product, Microsoft 365 Copilot.

Microsoft also confirmed that customers will soon be able to run the R1 model locally on the Copilot+ PC device. The move aims to address concerns regarding privacy and data sharing in the use of AI models.

However, challenges remain. DeepSee has stated that user data is stored on servers in China, which could be an obstacle in the adoption of this model in the United States.

Meanwhile, Microsoft and OpenAI are reportedly investigating allegations that OpenAI's technological data output was obtained illegally by parties linked to DeepSek. A report from Bloomberg News on Tuesday, January 28, revealed an investigation into the possible breach.

DeepSek's presence in the AI realm has prompted competitors to respond. OpenAI CEO Sam Altman stated that his company will accelerate the launch of the update, which will then be realized with the release of the ChatGPT version specifically designed for US government agencies.

Not only Microsoft and OpenAI, the Chinese technology giant, Alibaba, also released the latest version of the Qwen 2.5 AI model on Wednesday 29 January, coinciding with the first day of Chinese New Year celebrations.