Leaders in health informatics at NYU Langone told reporters they expect fully autonomous clinical AI—algorithms that operate without human oversight for routine tasks—within roughly five years, citing imminent use cases such as blood‑pressure medication titration and diabetic retinopathy screening. They framed autonomy as a response to workforce shortages and system inefficiencies rather than solely an efficiency play. NYU Langone highlighted validation pathways, workflow integration, and the need for robust safety governance as prerequisites for clinical deployment. The group argued some routine decision-making could be standardized enough to permit algorithmic execution with appropriate monitoring. Why it matters: hospital systems, regulators, and vendors must prepare for accelerated operational and regulatory questions around liability, monitoring, and validation as AI moves from assistive to autonomous roles in patient care.