Microsoft says its AI beats doctors at diagnosing, by four times

Microsoft says it has built an AI tool that can diagnose medical conditions up to four times more accurately than human doctors. The company published the findings Thursday in Science, the prestigious journal. The results come from tests that pitted Microsoft's AI system head-to-head against human physicians. The matchups used scenarios pulled from medical case reports and emergency department data.

Microsoft rounded up those cases and compiled them into its own benchmark for the face-off. That detail matters: the yardstick measuring the AI's performance was assembled by the same company making the claim. The company says the tool outperformed the doctors across the series of scenarios it tested.

Mustafa Suleyman, CEO of Microsoft AI, didn't hedge when he talked to The Verge. "It is finally here," he said, framing the moment as the payoff on years of promises that AI would transform medicine.

Healthcare has long been the showcase example AI leaders reach for when arguing their technology will improve lives. That vision has stayed mostly in the future tense, according to The Verge, until now. Microsoft is betting this study is the proof point that shifts the conversation.

The claim lands in one of the world's most prestigious journals, giving Microsoft a heavyweight platform for its pitch. Whether the broader medical community signs off on a tool graded against its maker's own benchmark is the fight ahead.