A program for renewing prescriptions, quietly launched in the U.S. state of Utah earlier this year, has sparked widespread medical debate: Is artificial intelligence ready to take on tasks that, until now, could only be performed by doctors? As Matthew Perron writes.

Online prescription renewals

The program allows Utah residents to avoid visiting a doctor's office and renew their prescriptions online using an AI-powered chatbot called 'Doctronic'.

On the surface, this step seems like a simple move toward making healthcare more convenient and accessible for patients and prescribing doctors. However, it also represents a significant milestone that breaks familiar rules, raising concerns and warnings among doctors, lawyers, and public health experts.

Doctronic's AI asks users to list their symptoms to create a treatment plan

Questions about AI's role in medicine

This pilot program has raised a set of questions about the role of AI in medicine, including how it should be regulated legally, whether doctors should be given the authority to challenge its decisions, and what safety measures are needed to protect patients.

The core of the debate lies in the fact that laws—both at the state and federal levels—restrict the authority to prescribe medications to licensed medical professionals. Supporters argue that these laws, which have been a cornerstone of American medicine for over 100 years, should be updated to include AI-powered chatbots and other modern technologies.

Dr. Eric Brisman from the University of Pennsylvania says: 'We have crossed a new threshold of granting a non-human entity a medical license, whether we want to call it that or not.'

AI cannot practice medicine under current laws

Brisman and other experts emphasize that they are not opposed to the idea of AI prescribing medications, but they believe strict standards comparable to those applied to human doctors—who undergo years of testing and training before obtaining a medical license—must be met.

Regulatory sandbox

In Utah, the Doctronic system was able to launch thanks to what is known as a 'regulatory sandbox,' which allows state officials to exempt promising AI companies from certain laws and restrictions.

The prescription renewal program is currently under the supervision of a five-member board of AI specialists—none of whom are doctors—who assert that they have implemented numerous safety measures. For example, during the initial phase of the program, human doctors review all renewal requests processed by the Doctronic system, while the supervising company expects to transition soon to a fully automated renewal system.

Chair of the medical licensing board unaware of the program's approval

The chair of the state's medical licensing board stated that he and his colleagues learned of the program when news of its launch was published in January. In a letter sent by 11 board members to state authorities in March, they called for a halt to the program, citing risks associated with automatic renewal of medications that may have side effects or cause drug interactions.

Doctors... no say

Dr. Alan Smith, a family physician who chairs the board (emphasizing that he was speaking in his personal capacity), said: 'We were basically told, yes, this is happening, and no, you don't have a say in it.'

Adding to the complexity is the fact that medical technology has traditionally been regulated at the federal level, while medical professions are regulated by states.

Doctronic officials view their AI technology as part of the practice of medicine, which is subject to state regulation. However, the federal Food and Drug Administration (FDA) is responsible for overseeing AI technologies that directly affect medical care or medical decision-making—a line that some experts believe Doctronic has crossed.

Some states pave the way for AI use in healthcare

In an interview with Doctronic officials to disclose whether they had sought approval from the FDA, Dr. Adam Oskowitz, who co-founded the company with a tech entrepreneur, said: 'Our goal here is simply to reach patients where they need healthcare. And we are trying not to get bogged down in the intricate details of the regulatory aspects.'

In Utah, residents can visit a website set up by Doctronic for the prescription renewal program. After identity verification, the AI chatbot asks users about their prescriptions and medical history, and checks for an existing valid prescription by linking to a national pharmacy database. If there are no issues, the AI can renew the prescription and send it to a local pharmacy. If further attention is needed, the chatbot transfers the patient to a doctor working within Doctronic's telehealth service.

Oskowitz envisions a future where many routine medical tasks—including ordering tests and analyzing results—could be delegated to Doctronic, allowing doctors to manage far more patients than they currently can.

Easing rules on AI use

Other states are also easing regulations on AI, including Texas and Wyoming.

Meanwhile, lawmakers in states such as Iowa, Idaho, and others have introduced legislation to grant formal licenses for AI-based medical services. Many of these bills are based on a model prepared by the Cicero Institute—a nonprofit think tank supportive of AI—founded by Joe Lonsdale, a co-founder of the AI software company Palantir. The institute's health policy director says opposition to AI use in medicine stems primarily from economic concerns among doctors and other healthcare workers.

Adam Meier, a policy director at the Cicero Institute, said: 'The first mover will face criticism and attack because there are economic interests, workforce concerns, and impacts on jobs.'

Doctors see potential risks in AI-driven prescription renewals

Smith, the medical board chair, says the risks to patients are real; he notes that the list of 190 medications eligible for renewal through the Doctronic system includes blood thinners, which could pose a danger if patients develop stomach ulcers or other conditions causing internal bleeding. Smith added: 'Often, when I see patients after six months, I find that their medical history or health status has changed. Just because a drug was prescribed before doesn't necessarily mean it's appropriate for the current situation.'