Microsoft expands its AI strategy by integrating DeepSeek R1 on Copilot+ PCs, while investigating its potential ties to OpenAI’s technology.
Microsoft now supports DeepSeek R1 AI on Windows 11 laptops, optimizing it for Snapdragon, Intel, and AMD processors. Learn more about the latest AI shift. Image: ChicHue |
Washington, USA — January 31, 2025:
In a notable move, Microsoft has unveiled plans to support the DeepSeek R1 reasoning model, making it available for on-device use, including on personal laptops. This follows Microsoft’s quick adaptation to the new AI technology, launching it first on its Azure cloud platform and GitHub for developers.
The company revealed that the DeepSeek AI model would come in NPU-optimized versions, designed specifically to work with Windows 11 Copilot+ PCs. These versions will be compatible with key processors like Qualcomm Snapdragon X devices, Intel Lunar Lake PCs, and AMD Ryzen AI 9 processors. The models will initially be available in smaller 7B and 14B versions, and later be integrated into Microsoft's AI Toolkit for developers.
“Optimized models allow developers to build and deploy AI applications directly on devices, leveraging the power of NPUs in Copilot+ PCs,” Microsoft explained in a blog post.
However, the support for DeepSeek raises some concerns. Microsoft’s push could signal a desire to reduce reliance on OpenAI, even as the company investigates potential intellectual property violations linked to DeepSeek’s training methods. DeepSeek, a China-based AI firm, has faced criticism for possibly using OpenAI’s code in its model, with allegations of distillation—a technique involving data extraction between models.
Despite privacy and legal concerns, Microsoft’s swift embrace of DeepSeek could provide users with more options and reinforce its Microsoft 365 Copilot product with diverse third-party models. The tech giant’s actions are viewed as a significant step in diversifying AI resources while addressing competition and legal challenges in the AI space.