Microsoft AI diagnostic system outperforms generalist doctors in complex case simulations

What's Happening

• Microsoft's AI diagnostic system diagnosed complex medical cases in a simulation more than twice as often as generalist physicians, according to a study published Thursday in Science.
• The MAI-DxO system coordinates several AI agents that act like a medical panel to question patients, order tests, and debate diagnoses.
• In a simulation using New England Journal of Medicine cases, the AI achieved 83.2% diagnostic accuracy compared to 36.1% for primary care doctors using reference materials.
• Microsoft tested the system across multiple models, but its best results came with OpenAI’s GPT-5 and GPT-5.5.

Why It Matters

Hospitals face a shortage of specialized diagnostic expertise. Multi-agent AI frameworks reduce diagnostic costs and improve accuracy on complex cases. This moves clinical AI beyond single-prompt chatbots into collaborative diagnostic panels.