Doctors are turning to AI for treatment decisions, not diagnoses

Clinical takeaway: Clinicians are rapidly incorporating AI-enabled information platforms into routine care, most often to support treatment and medication decisions rather than diagnostic questions.

AI tools are increasingly being integrated into clinical workflows, but until now there has been little real-world evidence on how clinicians actually use them. Understanding where and how physicians rely on AI can help inform evaluation, education, and governance of these technologies.

In this cross-sectional study published in JAMA, researchers analyzed use of a popular AI-enabled clinical information platform, among 952,008 US clinicians, including more than 556,000 physicians, from January 2024 through June 2026. Collectively, users submitted more than 271 million clinical queries during the 2.5-year study period.

Adoption was strikingly rapid. Monthly active users increased from just 355 clinicians in January 2024 to more than 703,000 by June 2026, while monthly query volume climbed to nearly 28 million. Physicians averaged 37 queries per month when active, while nurse practitioners and physician assistants averaged about 45 queries monthly.

Use was highest among trainees and early-career physicians. Doctors within 10 years of graduation generated nearly 62% of physician queries despite representing only about one-quarter of the US physician workforce. Oncology/hematology, internal medicine, and family medicine clinicians accounted for a disproportionately large share of AI use, whereas radiologists, orthopedists, ophthalmologists, and anesthesiologists used the platform less frequently.

Among 38.8 million classified queries, more than half focused on clinical decision-making. Treatment and management questions accounted for one-third of physician queries, making them the most common use case, while diagnosis represented just 13.9%. Physicians also frequently sought information outside their own specialty; 70% of surgeons' queries and 73% of anesthesiologists' queries involved nonspecialty topics.

“These findings highlight the growing role of AI-enabled clinical information platforms in clinical practice and suggest that efforts to evaluate, teach, and integrate AI in clinical medicine should be informed by empirical studies of its use in actual practice,” the authors wrote.

Source: Patel VR, et al. (2026 Oct 15) JAMA. Clinician Use of an Artificial Intelligence–Enabled Clinical Information Platform