Medical AI just got a report card. OpenEvidence has launched EvidenceGrade, a feature that scores and visualizes the quality of medical evidence behind every clinical answer it generates — think A through D letter grades based on study design, consistency, and relevance. The company also announced a system-wide rollout at NewYork-Presbyterian, Columbia, and Weill Cornell Medicine.
Medical AI just got a report card
OpenEvidence, an AI-powered medical search engine used by nearly 1 million clinicians, has launched EvidenceGrade — a feature that scores, grades, and visualizes the quality of published medical evidence behind every clinical answer it generates. Built on the widely adopted GRADE framework (the same methodology used by Cochrane and the WHO), the tool assigns letter grades (A through D) based on study design strength, consistency, precision, and how directly the evidence applies to the clinical question at hand.
The move reflects a broader shift in healthcare AI: companies are no longer just competing on speed or accuracy — they're competing on trust. By surfacing evidence quality in real time, OpenEvidence aims to help clinicians understand how much weight an AI-generated answer can actually bear in high-stakes decisions. On the same day, OpenEvidence announced a collaboration with NewYork-Presbyterian and its affiliated schools — Columbia University Vagelos College of Physicians and Surgeons and Weill Cornell Medicine — making the platform available across all hospitals and care sites.
Quick Facts
Why it matters: As AI becomes embedded in clinical workflows, knowing how good the underlying evidence is — not just what the answer is — could meaningfully reduce the risk of AI-informed medical errors.
