Microsoft MAI-DxO diagnostic system outperforms physicians in Science study

A peer-reviewed Science paper showed Microsoft’s MAI-DxO orchestrator, which coordinates multiple AI agents to ask questions and order tests sequentially, achieved 83% accuracy on complex New England Journal of Medicine cases versus 36% for generalist physicians using reference materials. On emergency-department cases the system scored 80% against 47% for physicians, often at lower estimated testing cost. Best results used OpenAI models; a prospective clinical study is planned.