Utah Becomes First US State to Let AI Renew Prescriptions Without a Doctor Visit

In Utah, any patient can now renew a prescription by talking to a chatbot alone. The program, called Doctronic, has been running since January 2026 and is the first system in the United States where artificial intelligence independently decides whether to extend a course of treatment, with no exam and no conversation with a licensed physician.

The mechanism is simple. A patient goes to the Doctronic website, confirms their identity, and the chatbot asks about current medications and medical history. The system checks a nationwide pharmacy database to verify the prescription actually existed and when it expires. If everything checks out, the AI sends a new prescription directly to the local pharmacy. In cases that raise doubts, the chatbot refers the patient to a physician working for Doctronic's telehealth platform.

A regulatory sandbox

The program was able to launch thanks to a so-called regulatory sandbox, a mechanism that lets state officials temporarily suspend rules for companies offering promising technology. The pilot is overseen by a five-person committee that does not include a single physician. The decision to approve Doctronic was made without consulting the state's medical licensing board, which learned about the program only after it had already started.

The board's chair, family physician Alan Smith, makes no secret of his frustration. In comments to the media, he said plainly that decisions are being made over the heads of the medical community. The board formally demanded the pilot be suspended in March, citing risks to patients on medications that require regular health monitoring. State officials refused, citing the benefits for patients with limited access to healthcare.

Yes, this is happening. And no, you don't have a say in it. - Alan Smith, chair of Utah's medical licensing board

Risk with blood thinners

The biggest controversy centers on the presence of blood thinners among the 190 approved medications. Experts note that their safe use depends on a patient's current health status, for example whether a stomach ulcer has developed that could lead to internal bleeding. Automatically renewing a prescription without a current exam can, in such cases, mask a serious risk that neither the patient nor an algorithm that only asks about reported symptoms would catch.

Physician Eric Bressman of the University of Pennsylvania described it as crossing a line medicine has never crossed before. The American Medical Association, for its part, warned that renewing a prescription should never be treated as a formality, since every such decision requires assessing whether the treatment is still appropriate for that particular patient.

We've crossed a threshold in terms of granting something that isn't a human the authority to practice medicine. - Eric Bressman, University of Pennsylvania

No federal oversight

At the federal level, no one is formally responsible for overseeing tools of this kind. The FDA has not approved any AI prescribing tool and has taken a wait-and-see approach to the whole phenomenon, issuing neither bans nor guidelines. Doctronic co-founder and physician Adam Oskowitz defends the program on the grounds of healthcare access, noting that for many patients the only alternative is no prescription at all.

Our goal is simply to reach patients wherever they need healthcare. - Adam Oskowitz, Doctronic co-founder

Politics also lies behind the push for such systems. Adam Meier, health policy director at the Cicero Institute, belongs to an organization founded by Joe Lonsdale, co-founder of Palantir, which promotes deregulation as a way to cut US healthcare costs. Daniel Aaron, a law professor at the University of Utah, points out that regulatory sandboxes, originally designed for financial technology, lack clear accountability mechanisms for when a system error affects health rather than money.

For observers elsewhere, the Doctronic case is a signal of how quickly automation can outpace regulation even in a system as heavily regulated as US healthcare. Similar programs are not yet operating in Poland, but work is underway on an EU regulation covering AI-based medical devices, and the debate over liability for errors by medical algorithms is only just getting started. The Utah case raises questions that will sooner or later land on the desks of European regulators too: who is responsible when a chatbot makes a wrong call, and does access to treatment justify giving up physician oversight.

The Utah program continues to operate for now despite the medical board's objections, and other states are watching its progress rather than waiting for federal rules. If the pilot survives the coming months without a high-profile incident, more states are likely to begin testing similar systems, and the issue of automated prescribing could end up before Congress as well.

Sources: Is AI ready to take over your prescriptions? Doctors are wary of Utah's automated refill program (washingtontimes.com), Doctors sound alarm as AI starts renewing prescriptions without a checkup (washingtontimes.com)