A human in the loop is only oversight when the decision is informed, proportionate to the stakes, and free to be "no." New essay on agentic AI in healthcare.
"That's not human judgment. That's automation bias with a human signature attached." Anthropic's Claude said that about armed conflict. It applies with equal force to medicine, where a clinician glancing at a queue of flagged studies is not really deciding anything. This essay lays out a framework for building agentic AI in healthcare that the patient, the clinician, and the administrative staff whose workflows the agent touches can actually endorse. Three commitments. Three tests. One governance-agent pattern OpenCoreEMR is building in the open.