83%: The Accuracy Score That Could Reshape Medical Diagnosis

A new AI diagnostic system has outperformed human doctors in a rigorous simulation of complex medical cases.

Microsoft AI published a study in the journal Science on Thursday, revealing that its system, known as MAI-DxO or simply DxO, correctly diagnosed intricate cases 83% of the time. By comparison, primary care doctors and general internists working alone with reference materials achieved a 36% success rate. The system functions by coordinating multiple AI agents that act like a medical panel, questioning virtual patients, ordering tests, and debating potential diagnoses before reaching a conclusion.

The study utilized complex cases drawn from the New England Journal of Medicine, a dataset typically handled by specialist teams rather than generalists. Despite this distinction, Microsoft AI CEO Mustafa Suleyman described the results as a significant step toward achieving what he calls medical superintelligence. Dominic King, vice president of Health at Microsoft AI, noted that the current healthcare system faces a growing mismatch between the demand for expertise and its supply. He stated, "There is something very substantial and exciting here," as Microsoft joins Google and OpenAI in claiming that its artificial intelligence can match or exceed human performance.

Analysis of the data suggests that much of the improvement stems from more powerful underlying models, with Microsoft's orchestration layer providing additional gains. The team tested the system using models from OpenAI, Google, Anthropic, Meta, and DeepSeek. The highest accuracy was achieved using OpenAI's GPT-5 and GPT-5.5 models. Cost analysis showed that GPT-5 alone scored 78.9% on complex cases at a cost of approximately $8,400 per case. While DxO reduced that cost by about $3,000, it only increased accuracy by roughly four points. On emergency department cases, the system offered little improvement in either cost or accuracy compared to the standalone model.

With the newer GPT-5.5 model, released in April 2026, the system increased accuracy on complex cases from 72% to nearly 87% at a slightly lower cost. However, these gains were less pronounced for weaker models, such as Gemini 2.5 Flash, which jumped from 46.4% to nearly 69% with the assistance of DxO. Researchers noted that accuracy gains naturally shrink as models approach theoretical maximums. The orchestration system is designed to optimize both accuracy and cost across the broader challenge of clinical decision support.

Microsoft is currently partnering with the Mayo Clinic to train a dedicated healthcare foundation model, with a release planned for next year. Suleyman believes this infrastructure will enable enormous leaps in the field. He also announced a new partnership with Beth Israel Deaconess Medical Center in Boston to evaluate DxO in real-world hospital settings. This study will compare the AI's recommendations against human doctors in both emergency and urgent care environments. King described the initiative as potentially the largest study of its kind, noting that the technology is ready for real-world evaluation.

Alongside the study, Microsoft published a manifesto defining medical superintelligence as the capacity to surpass contemporary human-led health systems in reasoning and coordination. Suleyman, who previously co-founded DeepMind, aims to make the company the leader in clinical decision support products. He expects healthcare models to become an order of magnitude better in capability within the next 24 to 36 months, driven by a flywheel of real-world usage and iteration.