Microsoft AI Diagnostic Tool Beats Doctors With 83% Accuracy on Complex Medical Cases
12 Sources
[1]
Microsoft's Mustafa Suleyman says its AI outperforms doctors on hard cases as he chases 'medical superintelligence'
Microsoft's AI diagnostic system correctly diagnosed complex medical cases in a simulation more than twice as often as generalist physicians, according to a study published Thursday in Science. Mustafa Suleyman, Microsoft AI's CEO, called the result a step toward "medical superintelligence." The
[2]
Microsoft says its AI tool is better than doctors at diagnosing diseases
For years, AI leaders have touted healthcare as one of the clearest examples of where their technology will improve lives. Now "it is finally here," Microsoft AI CEO Mustafa Suleyman tells The Verge. The company claims to have developed an AI-powered tool it says can diagnose conditions up to four
[3]
Microsoft Develops AI Tool That Can Diagnose 'sequentially' Like a Doctor
Oct 22 (Reuters) - Microsoft researchers say an AI tool that diagnoses patients step by step, as doctors do, was more than twice as accurate as general physicians at solving complex medical cases, according to a study published on Thursday in Science. The model identifies and coordinates the
[4]
Microsoft develops AI tool that can diagnose 'sequentially' like a doctor
Oct 22 (Reuters) - Microsoft (MSFT.O), opens new tab researchers say an AI tool that diagnoses patients step by step, as doctors do, was more than twice as accurate as general physicians at solving complex medical cases, according to a study published on Thursday in Science, opens new tab. The
[5]
Microsoft AI tool beat generalist doctors at step-by-step diagnosis
A Science study found Microsoft's orchestrator more than twice as accurate as generalist physicians on 304 complex case reports. Microsoft has developed an AI diagnostic tool that solved complex medical cases more than twice as accurately as generalist doctors in a study published in
[6]
Microsoft Says Its New AI Is More Than Twice as Good at Diagnosing Diseases as Doctors
Microsoft plans to start testing the tool in a Boston hospital soon. Microsoft says its medical AI system can diagnose difficult cases more than twice as accurately as practicing doctors. Dubbed the Microsoft AI Diagnostic Orchestrator (MAI-DxO), the system "substantially outperformed practicing
[7]
Microsoft Says Its New AI Is Better at Diagnosing Patients Than Real Doctors
Microsoft's new AI is surpassing the abilities of human doctors -- according to its own study, at least. The research, which was peer-reviewed and published in the prestigious journal Science, showed that the medical model, dubbed the Microsoft AI Diagnostic Orchestrator (MAI-DxO), was over twice
[8]
Microsoft AI: Microsoft develops AI tool that can diagnose 'sequentially' like a doctor
Microsoft researchers say an AI tool that diagnoses patients step by step, as doctors do, was more than twice as accurate as general physicians at solving complex medical cases, according to a study published on Thursday in Science. The model identifies and coordinates the questions that should be
[9]
Microsoft AI Tool Diagnoses Patients Like Doctors, Study Finds New Hope
Microsoft has developed an AI tool that diagnoses patients step by step. Researchers found it more than twice as accurate as general physicians in complex cases, with further clinical testing planned in Boston. Microsoft researchers have developed an artificial intelligence tool that diagnoses
[10]
Microsoft publishes Science study: its AI diagnostic AI outperforms doctors
Microsoft published a Science paper today on MAI-DxO, its AI diagnostic system, with new results on how the software handled simulated medical diagnosis. Across 304 difficult New England Journal of Medicine cases, MAI-DxO reached 83.2% accuracy, according to Microsoft's paper, compared with 36.1%
[11]
Microsoft develops AI tool that can diagnose 'sequentially' like a doctor
Oct 22 (Reuters) - Microsoft researchers say an AI tool that diagnoses patients step by step, as doctors do, was more than twice as accurate as general physicians at solving complex medical cases, according to a study published on Thursday in Science. The model identifies and coordinates the
[12]
Can AI beat doctors at diagnosing medical conditions? Microsoft claims its AI can do it at lower costs, here is how
The technology reportedly reduced diagnostic costs, but its effectiveness in real-world healthcare settings remains unproven. AI use cases in healthcare have been an aspirational vision in the medical industry, and it appears Microsoft is working in this direction. In a recent conversation with
Share
Copy Link
Microsoft AI CEO Mustafa Suleyman unveiled an AI diagnostic tool that outperformed generalist physicians by more than twice on complex medical cases. The MAI Diagnostic Orchestrator achieved 83% accuracy compared to 36% for human doctors in a Science journal study. The system works step-by-step like a medical panel, questioning patients and ordering tests before reaching diagnoses.
Microsoft AI has developed a diagnostic tool that achieved 83% accuracy on complex medical cases, more than double the 36% accuracy rate of generalist physicians working with reference materials, according to a study published in Science journal
1
2
. The system, called MAI Diagnostic Orchestrator (MAI-DxO), represents what Microsoft AI CEO Mustafa Suleyman describes as a step toward "medical superintelligence."1
How the AI Diagnostic Tool Works Through Step-by-Step Diagnosis
Unlike traditional medical AI systems that analyze all patient information simultaneously, the MAI Diagnostic Orchestrator mimics how doctors actually work
3
4
. The AI in healthcare system coordinates multiple AI agents that function like a panel of clinical experts, questioning patients, ordering tests, and debating diagnoses before reaching conclusions1
. The orchestrator first decides what questions to ask about patient history and physical examination findings. Based on initial information, it determines which tests to order, reviews results, and requests additional testing as needed while continuously updating working hypotheses3
. This step-by-step diagnosis approach marks a shift from earlier AI systems that relied on multiple-choice assessments. "No patient ever came to see me with a nice summary," said Dr. Dominic King, corporate vice president at Microsoft AI Health4
.Test Results Show AI Outperforms Doctors on Complex Cases
The study tested the AI diagnostic tool on over 300 highly unusual medical dilemmas from the New England Journal of Medicine case reports
3
5
. On these diagnosing complex medical cases, MAI-DxO achieved 83.2% accuracy while generalist physicians using textbooks and internet resources scored only 36.1%5
. The system also demonstrated strong performance on emergency department cases drawn from Beth Israel Deaconess Medical Center admissions data, scoring 80.3% accuracy compared to 46.8% for human physicians2
5
. Microsoft AI measured both accuracy and estimated cost per case. Running on OpenAI GPT-5, the system processed complex cases at approximately $5,400 per diagnosis compared to $8,400 when GPT-5 operated independently without the orchestrator1
. "We're not just measuring the accuracy of the systems. We're also measuring how judiciously and carefully it goes through a diagnostic work-up," explained Harsha Nori, head of AI at Microsoft AI Health4
.Performance Varies Based on Underlying AI Models
Microsoft tested the MAI Diagnostic Orchestrator with models from OpenAI, Google, Anthropic, Meta, and DeepSeek, with best results coming from OpenAI GPT-5 and GPT-5.5
1
. The data reveals that much of the improvement stems from more powerful base models, with the orchestration system adding gains on top. For medical diagnosis accuracy, GPT-5 alone scored 78.9% on complex cases, while the orchestrator boosted this to 83%1
. With GPT-5.5, released in April 2026, the orchestrator increased accuracy from 72% to nearly 87% on complex cases1
. The orchestrator proved most valuable when working with less capable models. Google's Gemini 2.5 Flash scored 46.4% independently but reached nearly 69% with orchestration support1
. However, on emergency department cases, the orchestrator provided minimal improvement regardless of the underlying model.Related Stories
Real-World Testing and Clinical Decision Support Plans
Microsoft plans to evaluate the system in actual hospital settings through a partnership with Beth Israel Deaconess Medical Center in Boston
1
4
. The initial focus will test whether the AI diagnostic tool can effectively triage patients in emergency department waiting rooms3
5
. King indicated testing will "get underway in the next couple of months"2
. The current study results come with limitations. The comparison involved generalist physicians rather than specialists or physician teams who typically handle such complex cases1
. Cases were presented as text without access to physical examinations or medical imaging5
. "It needs clinical validation," Suleyman acknowledged. "That is a leap, but this is a major milestone along that journey"2
. Microsoft compiled test cases into a benchmark called SDBench, which it plans to open-source2
.Microsoft's Vision for Medical Superintelligence
Mustafa Suleyman, who previously co-founded DeepMind before joining Microsoft AI in 2024, has spent his career applying AI to medical challenges
1
. He predicts healthcare AI models will improve by an order of magnitude within 24 to 36 months as real-world usage creates a feedback loop for identifying and patching gaps1
. Microsoft is collaborating with Mayo Clinic to develop a healthcare foundation model scheduled for release next year1
. "I really believe that in 2027, we will be on the verge of a coding style breakthrough for medical records," Suleyman stated. "I think we're going to be able to stream highly accurate predictions about how a medical record might unfold, and obviously that kind of foresight for a clinician is gold dust"1
. The company outlined five additional areas beyond diagnosis where AI must excel, including coordinating clinical care, making treatment decisions, and handling the human aspects of medicine2
. While Microsoft has not committed to commercialization timelines, Suleyman expects "extremely impressive deployments of intelligent systems in production" next year that will "change the way care is delivered in a big way"2
. Despite pursuing systems that could match human performance on medical superintelligence within years, Suleyman insists human clinicians will retain a central role, with AI ideally removing "the drudgery" from medicine rather than adding workload2
.References
Summarized by
Navi
[1]
Related Stories
Microsoft's AI Diagnostic Tool Outperforms Human Doctors in Complex Medical Cases
30 Jun 2025•Technology
AI Diagnosis Outperforms Emergency Room Doctors in Harvard Study, Raises Questions on Patient Care
30 Apr 2026•Science and Research
AI Outperforms Physicians in Virtual Urgent Care Study, Highlighting Potential for Improved Patient Care
05 Apr 2025•Health
Recent Highlights
Recent Highlights
Today's Top Stories
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.
