For decades, medicine has relied on a simple safeguard: when it comes to prescribing medications, a human clinician is always in the loop. That assumption is now being tested in a first-of-its-kind pilot. In Utah, regulators have approved the use of a generative AI system to monitor patients with chronic disease and automatically renew medications based on the individual’s health status — all without human oversight.

Last month’s poll asked where you stood on this development. Here are the results:

My thoughts: The results of the survey demonstrate how far people have come in accepting generative AI in medical care … and how wide the trust gap remains. Nearly half of respondents support Utah’s pilot outright or with some reservations. Only 11% oppose it entirely. That suggests readers are increasingly open to generative AI playing a direct role in chronic disease management.

But the second question draws a clear boundary: Even if outcomes equal or exceed those of human clinicians, 71% of respondents oppose allowing AI to operate without human oversight. Only 28% support full autonomy.

This pattern reflects the familiar trajectory of technological adoption: initial skepticism, cautious acceptance, then gradual expansion as evidence accumulates. The remaining question is to what extent autonomy will be granted. ATMs and travel websites operate without human oversight. Robinhood, the online trading app, does as well. But brokerage houses still require a human conversation. Medicine, along with regulators, will need to decide where to land on the continuum.