Microsoft claims its AI can diagnose patients better than doctors
What's the story
Microsoft says its artificial intelligence (AI) tool can diagnose medical conditions with up to four times the accuracy of human doctors. The company's AI CEO, Mustafa Suleyman, shared the news with The Verge. The findings were published in the prestigious journal Science and involved tests comparing Microsoft's AI system against human physicians on various scenarios derived from medical case reports and emergency department data.
Performance comparison
AI system outperformed doctors in diagnostic accuracy
The tests showed that Microsoft's AI system, Diagnostic Orchestrator, achieved a diagnostic accuracy of around 83% on cases taken from hundreds of reports published in The New England Journal of Medicine.
In contrast, generalist physicians using references had an accuracy of about 36%, while those working without them scored only 22%.
The AI system also outperformed doctors on emergency department cases drawn from a large database of patient admissions at Boston's Beth Israel Deaconess Medical Center.
Cost efficiency
Microsoft's AI system estimated to be cheaper than human doctors
Not only did Microsoft's AI system outperform human doctors in terms of diagnostic accuracy, but it also did so at a lower estimated cost.
The researchers behind the study noted that this system, which is powered by OpenAI's GPT-5, outperformed frontier AI models operating independently.
However, they didn't include newer models like GPT-6 Astra, Meta's Muse Wonder Full, Google's Gemini 4 Argon, or Anthropic's Opus 5.5 in their peer-reviewed paper.
Validation need
Need for clinical validation before widespread use
Despite the impressive results, Microsoft acknowledges that there is still a long way to go before this technology can be widely deployed.
The system has only been tested on past medical cases and its performance in real clinical settings remains uncertain.
Suleyman emphasized that "it needs clinical validation," adding that while this is a major milestone, "that is a leap" toward broader application of the technology.
Testing phase
Testing the AI system in real-world clinical environments
Microsoft plans to test the system in hospital settings with Beth Israel Deaconess Medical Center. This will help assess its performance in real-world clinical environments.
However, neither Dr. Dominic King, a co-author of the study, nor Suleyman have committed to a firm timeline for this testing phase.
The company has also not revealed when or how the technology might become available commercially.
Future impact
"Next year we're going to see extremely impressive deployments": Suleyman
Suleyman is optimistic about the future of intelligent systems in production, saying "next year we're going to see extremely impressive deployments." He believes this technology will greatly change how care is delivered.
However, he also stressed that human clinicians should always play a central role in patient care.
The company has already established a strong foothold in US healthcare with its digital tools and Copilot chatbot.