Microsoft’s MAI-DxO multi-agent diagnostic system correctly diagnosed complex medical cases 83% of the time in a simulation published in Science, compared with 36% for primary care doctors and general internists working alone with reference materials. The system coordinates AI agents that question a patient, order tests, and debate a diagnosis, using complex New England Journal of Medicine cases. Specialists and physician teams were not part of the comparison. Best results used OpenAI’s GPT-5 and GPT-5.5; gains from DxO orchestration were larger on weaker models and complex cases than on emergency department cases. Microsoft AI CEO Mustafa Suleyman called the work a step toward “medical superintelligence.” Microsoft is working with the Mayo Clinic on a healthcare foundation model planned for next year and partnering with Beth Israel Deaconess Medical Center to evaluate DxO in real hospital settings.
Microsoft’s MAI-DxO multi-agent diagnostic system correctly diagnosed complex medical cases 83% of the time in a simulation published in Science, compared with 36% for primary care doctors and general internists working alone with reference materials. The system coordinates AI agents that question a patient, order tests, and debate a diagnosis, using complex New England Journal of Medicine cases. Specialists and physician teams were not part of the comparison. Best results used OpenAI’s GPT-5 and GPT-5.5; gains from DxO orchestration were larger on weaker models and complex cases than on emergency department cases. Microsoft AI CEO Mustafa Suleyman called the work a step toward “medical superintelligence.” Microsoft is working with the Mayo Clinic on a healthcare foundation model planned for next year and partnering with Beth Israel Deaconess Medical Center to evaluate DxO in real hospital settings.