According to a study published in *Science* magazine, Microsoft's AI diagnostic system, MAI-DxO, achieved an 83% accuracy rate in simulating the diagnosis of complex cases, compared to 36% for general practitioners (using reference materials). Mustafa Suleyman, CEO of Microsoft AI, called this a step towards "medical superintelligence." The system coordinates multiple AI agents, asking questions, issuing orders, and discussing diagnoses like a team of doctors. The study found that most of the performance improvement came from more powerful underlying AI models, with Microsoft's orchestration system providing additional gains. Microsoft is collaborating with the Mayo Clinic to train a foundational model for healthcare, slated for release next year, and will work with Beth Israel Deaconess Medical Center in Boston to evaluate DxO's clinical reasoning capabilities in various hospital settings.