OpenEvidence, the AI platform used widely by practicing US physicians, announced a strategic collaboration with Johns Hopkins Medicine on October 13 to redesign both medical education and clinical decision support for an AI-saturated era of practicing medicine. The partnership follows a similar deal the company struck with Penn Medicine in September, extending OpenEvidence’s reach into two of the country’s most prominent academic medical systems within a single month.
Two tracks: training doctors, and helping them decide
The collaboration splits into distinct pieces. On clinical care, OpenEvidence and Johns Hopkins plan to build tools that combine published medical evidence with an individual patient’s history and the hospital system’s own practice standards — moving beyond generic clinical decision support toward recommendations tailored to how Hopkins specifically treats conditions. Sara Murray, Johns Hopkins’ chief medical informatics and AI officer, said the goal is to identify high-impact clinical use cases, build tools that fit naturally into existing clinician workflows, and rigorously evaluate their effects on decisions, quality, and patient outcomes. On education, the plan centers on an adaptive AI-based training model built around foundational skills: pathophysiology, evidence-based medicine, clinical reasoning, diagnosis, and patient communication.
Why academic medicine is moving now
The deal reflects a broader reckoning inside medical schools about what it means to train a doctor when AI tools can already answer many of the factual-recall questions that have anchored medical education for a century. Rather than banning AI tools or treating them as a threat to clinical judgment, Johns Hopkins is betting that the more durable skill is knowing how to work alongside them — a philosophy that mirrors how other professional fields, from law to software engineering, have begun restructuring training around AI-assisted workflows rather than against them.
OpenEvidence’s rapid institutional expansion
The Hopkins deal is the latest in a string of moves establishing OpenEvidence as the default AI layer inside major US health systems. In September, the company announced a global partnership with Penn Medicine that will bring AI-powered clinical decision support to more than 10,000 clinicians across Penn’s seven-hospital network. OpenEvidence has also said it closed $250 million in funding in January at a $3.5 billion valuation, and has touted a single-day milestone of 1 million clinical consultations between verified doctors and its AI system — a scale claim that, if accurate, would put daily usage well ahead of most competing clinical AI tools.
The caution from outside observers
Both organizations acknowledge they are still early in defining exactly what capabilities the partnership will produce, and independent researchers note that none of OpenEvidence’s clinical tools have yet been subject to the kind of peer-reviewed, randomized evaluation that would establish whether AI-assisted decisions actually improve patient outcomes rather than simply speeding up the process of reaching a decision a physician would have made anyway. There is also a structural concern raised by health-policy researchers: as a handful of AI vendors become embedded across multiple major health systems simultaneously, any flaw or bias in their underlying models could propagate across institutions that currently operate independently.
What’s next
Johns Hopkins and OpenEvidence say they are still in the early stages of defining specific tools and curriculum changes, with no public timeline yet for when AI-assisted training modules or new clinical decision-support features will reach students and physicians. The partnership’s progress will likely be watched as a bellwether for how quickly other academic medical centers follow Hopkins and Penn in formalizing relationships with a single AI vendor rather than building internal tools independently.
Photo: Polina Tankilevitch / PEXELS via Pexels
