Microsoft’s AI Diagnostic Tool Beats Human Doctors in Accuracy Tests
Tap title to read story
Microsoft’s new AI system, the Diagnostic Orchestrator, achieved diagnostic accuracy of about 83 % on cases from The New England Journal of Medicine, compared with roughly 36 % for generalist physicians using references and 22 % for those working without them. On emergency‑department cases from Beth Israel Deaconess Medical Center, the AI scored 80 % accuracy versus an average of 47 % for human physicians. The system, powered by OpenAI’s GPT‑5 and built on Microsoft’s SDBench benchmark, also reached diagnoses at a lower estimated cost than human counterparts. The findings were published Thursday in Science. Microsoft says the technology still requires further development before wide deployment.
More context
Timeline
- 2020-01 — Microsoft launches AI for Health Program
- On June 30 2025 Microsoft publishes Diagnostic Orchestrator study in Science.
- On June 30 2025 Diagnostic Orchestrator achieves eighty‑three percent accuracy on case reports.
- On June 30 2025 System outperforms doctors on emergency department data.
- On June 30 2025 Microsoft announces need for clinical validation.
Sources
JQJO Editorial Desk · Human reviewed · AI assisted. Report a correction.
