The US tech giant announced the move on Wednesday, listing R1 in its model catalog alongside more than 1,800 other AI models.
DeepSeek launched its AI assistant last week, positioning it as a cost-effective alternative to existing services. By Monday, it had overtaken OpenAI’s ChatGPT as the most downloaded app on Apple’s App Store, rattling tech investors and fueling concerns about the competitive landscape.
Microsoft’s latest move corresponds with its strategy to diversify beyond OpenAI, its key AI partner. The company has been integrating internal and third-party models to power Microsoft 365 Copilot, a shift Reuters reported last month. Additionally, Microsoft plans to allow users to run the R1 model locally on Copilot+ PCs, which could help address privacy and data-sharing concerns.
However, DeepSeek’s data storage practices may pose a challenge for its US adoption. The company has stated that it stores user information on servers in China, a detail that could raise regulatory scrutiny.
Meanwhile, Microsoft and OpenAI are reportedly investigating whether a group linked to DeepSeek obtained OpenAI’s proprietary data without authorization, according to Bloomberg News.
DeepSeek’s rapid ascent has intensified competition, prompting responses from industry leaders. OpenAI CEO Sam Altman said the company would “pull up some releases” and followed through by launching a government-tailored version of ChatGPT for US agencies on Tuesday.
China’s Alibaba also entered the fray, unveiling an upgraded Qwen 2.5 AI model on Wednesday—an unusual move as it coincided with the start of the Lunar New Year.
