Microsoft AI Outperforms Doctors in Complex Diagnoses by 47 Points

An artificial intelligence system developed by Microsoft correctly diagnosed complex medical cases in a simulation more than twice as often as human generalist physicians. The disparity in performance highlights a significant shift in how technology might approach difficult clinical reasoning, offering a potential solution to the growing gap between patient demand and available medical expertise.

The system, known as MAI-DxO or DxO for short, functions by coordinating multiple AI agents that act like a panel of physicians. These agents question virtual patients, order tests, and debate potential diagnoses. In a study published this week in Science, the system used complex cases from the New England Journal of Medicine and achieved an accuracy rate of 83%. By comparison, primary care doctors and general internists working individually with reference materials were correct only 36% of the time. It is important to note that such complex cases are typically handled by specialists or teams in real-world settings, who were not included in this specific comparison.

Dominic King, vice president of Health at Microsoft AI, described the results as substantial and exciting. He emphasized that the current biggest challenge in healthcare delivery is the mismatch between the supply of medical expertise and the rising demand. By publishing these findings, Microsoft joins Google and OpenAI in asserting that its AI tools can match or exceed the capabilities of human doctors in specific diagnostic tasks.

The study indicates that much of the performance improvement stems from the use of powerful underlying AI models, with Microsoft’s orchestration system providing additional gains. Researchers tested the system using models from several companies, including OpenAI, Google, Anthropic, Meta, and DeepSeek. The best results were achieved using OpenAI’s GPT-5 and GPT-5.5 models. When using GPT-5 alone with a basic prompt, the system scored 78.9% on complex cases at an estimated cost of $8,400 per case. Adding the DxO coordination reduced that cost by approximately $3,000 while raising accuracy by about four points. However, on emergency department cases, the improvements in both accuracy and cost were minimal between the standalone model and the assisted system.

Mustafa Suleyman, CEO of Microsoft AI, stated that this result is a step toward what he calls medical superintelligence. He is working with the Mayo Clinic to train a foundation model for healthcare, with a release planned for next year. Suleyman believes this effort will allow the company to make significant leaps in the field. He expressed confidence that by 2027, the company will be on the verge of a breakthrough in coding medical records, enabling highly accurate predictions about how a patient’s record might unfold. He noted that such foresight is incredibly valuable for clinicians.

The stakes for the industry are high as Microsoft seeks to integrate these tools into broader healthcare systems. Through a new partnership with Beth Israel Deaconess Medical Center in Boston, the company will evaluate the clinical reasoning capabilities of DxO across different hospital settings. This study will compare AI recommendations against those of human doctors in parallel. King noted that the technology is ready for real-world evaluations, particularly in complex environments like a Boston emergency department. Microsoft aims to have the state-of-the-art medical clinical decision support products over the next few years, building on its existing dictation and radiology tools.

Suleyman, who previously co-founded DeepMind, has spent his career using AI to solve medical issues. He expects healthcare models to become an order of magnitude better in capability within the next 24 to 36 months. He compared the current phase to the early days of chat and coding AI, where progress accelerates once a flywheel of real-world usage and iteration begins. He believes this cycle will soon start for medical records and clinical decision support, marking a major turning point for the industry.