Thank you so much for your interest in our event. Be on the lookout for more Policy Events soon!
About the Event
The American Medical Informatics Association (AMIA) hosted an in-person policy forum exploring Human-in-the-Loop (HITL) and Human Oversight as approaches to clinical AI oversight and governance that strengthen safety, accountability, and trust while supporting innovation. The event convened AMIA members, federal agency leaders and Congressional staff, and patients to examine how human judgment can be embedded by design across the AI lifecycle, from model development to real-world clinical use.
Why Defining the Role of Humans Matters
Over the past several years, humans have increasingly relied on AI tools in our everyday lives. Given the national clinician shortages and increasing clinical documentation times, both clinicians and patients have much to gain from the potential of AI to make care more efficient and accessible, allowing the use of AI in healthcare contexts mainstream. AMIA’s position on AI in clinical decision support (CDS) is that its use is neither good nor bad, it’s simply neutral. AI is a tool that can be used appropriately, supporting provision of and access to care, or it can be a tool that is used poorly, resulting in unintended consequences due to misuse, badly trained models, performance degradation, and distrust in the models. Human oversight accounts for the humanity needed in patient care. AI can optimize prediction, but without human involvement, healthcare loses its primary safety mechanism: contextual judgment under uncertainty.

Key Policy Questions