Microsoft is shaking things up in the AI world by adding DeepSeek’s R1 model to its Azure AI Foundry service. This platform brings together a variety of AI tools for businesses, and R1 is the latest addition. Microsoft promises that the model has been thoroughly tested for safety and reliability, ensuring it’s ready for real-world use.

But that’s not all. The tech giant hinted that “distilled” versions of R1 will soon be available for Copilot+ PCs, Microsoft’s AI-ready Windows devices. This means users could soon run advanced AI models directly on their computers, no internet required.

The move is somewhat surprising, given recent reports that Microsoft and OpenAI are investigating DeepSeek for possibly misusing their services. Despite this, Microsoft seems eager to embrace R1’s potential. Experts speculate that the model’s capabilities might have been too impressive to ignore.

However, R1 isn’t without its flaws. Tests by NewsGuard revealed that it struggles with accuracy, often providing incorrect or vague answers to news-related questions. Additionally, it tends to avoid responding to prompts involving China, likely due to local censorship rules.

Whether Microsoft tweaked the model to address these issues remains unclear. For now, the tech world is watching closely to see how businesses and developers put R1 to work. Could this be the next big leap in AI innovation? Only time will tell.