This change transforms genAI from an experimental capability into a verifiable system of record. The most forward-thinking health systems already maintain AI registries (similar to software bills of material) listing approved models, data sources, and governance owners. By next year, this practice will be standardized or well on its way.
What it looks like in the real world
Consider the challenge of caring for a patient with chronic illnesses such as diabetes and heart failure. Their data spans years of lab results, imaging, prescriptions, and clinical notes scattered across multiple EHRs. The old approach was to transfer the entire file into an LLM and ask, “What should happen next?” »
A modular, multi-agent approach works differently. An extraction agent structures the patient’s story, a reasoning agent identifies risk patterns, a drug review agent flags contraindications, and a conversational agent explains the results to clinicians in plain language. A governance layer follows each inference, ensuring transparency and auditability.