Utah has become the first state in the U.S. to allow artificial intelligence (AI)—without direct physician involvement—to renew certain medical prescriptions. The move, launched quietly through a pilot program with health-tech startup Doctronic, tests how far patients, regulators, and health systems are willing to trust AI with one of medicine’s most sensitive responsibilities: prescribing medication.

The initiative focuses on routine prescription renewals for patients with chronic conditions and represents an early real-world experiment in automating clinical decision-making at scale.

Patients using the system begin by visiting a webpage that verifies they are physically located in Utah. Once verified, the AI pulls the patient’s prescription history and presents a list of medications eligible for renewal.

The AI system then walks patients through the same structured clinical questions a physician would typically ask to determine whether a refill is appropriate. If the AI clears the renewal, the prescription is sent directly to a pharmacy.

The program is limited to 190 commonly prescribed medications. Certain drug categories—including pain management medications, ADHD drugs, and injectables—are excluded for safety reasons.

As an added safeguard, human physicians will review the first 250 prescriptions issued within each medication category to validate the AI’s performance. After that threshold is met, subsequent renewals in that category may be handled autonomously unless the system flags uncertainty and escalates the case to a doctor.

Doctronic also operates a nationwide telehealth practice that routes patients to human physicians following an AI-led consultation, providing an alternative pathway for cases that require direct medical oversight.

Matt Pavelle, Doctronic’s co-founder and co-CEO, said the goal is to reduce barriers to ongoing treatment for patients managing chronic conditions. “It’s hard to get a renewal—if you have a chronic condition and you can’t get your medication, terrible things happen,” he said.

State officials say the pilot addresses growing pressure on the healthcare system—particularly in rural areas—where clinicians are overextended and access to care is limited.

“Health care expenses keep climbing and clinicians are stretched thin,” said Margaret Busse, executive director of the Utah Department of Commerce. Automating routine prescription renewals, she said, could reduce medication lapses while easing the burden on providers and lowering patient costs.

Utah officials also view the program as a controlled way to support innovation in areas where technology is advancing faster than regulation.

Medical organizations caution that removing physicians from prescribing decisions introduces new risks.

In a statement, Dr. John Whyte, CEO and executive vice president of the American Medical Association (AMA), warned that while AI holds promise, “While AI has limitless opportunity to transform medicine for the better, without physician input it also poses serious risks to patients and physicians alike.”

Concerns include potential misuse or abuse—such as people struggling with addiction attempting to game automated systems to obtain drugs—as well as missed drug interactions or subtle clinical red flags that trained clinicians might otherwise catch.

Al Carter, CEO and executive director of the National Association of Boards of Pharmacy, noted that AI is already used in pharmacy workflows but questioned how regulators can evaluate whether rapidly evolving technology is safe and effective for patient care.

According to data shared with Utah regulators, Doctronic compared its AI system to human clinicians across 500 urgent care cases, reporting a 99.2% match rate in treatment decisions.

Co-founder Dr. Adam Oskowitz, an associate professor of surgery at the University of California, San Francisco, said the system is designed to err on the side of caution, automatically escalating cases to a physician when uncertainty arises. He pointed to built-in human oversight mechanisms as part of the company’s safety approach as the system transitions toward greater autonomy.

Doctronic has also secured a malpractice insurance policy specifically covering an AI system—treating the technology as accountable at the same level as a human physician.

“In medicine, there’s always going to be potential issues that patients have,” said Oskowitz. “Whether it’s caused by the AI or not — we will take the risk. I think this is going to be infinitely safer than a human doctor.”

The company currently charges $4 per prescription renewal, a price it describes as temporary. Executives say costs could drop as the system scales, with long-term plans to integrate renewals into insurance coverage or a low annual fee.

Doctronic is already in discussions with other states, including Texas, Arizona, and Missouri, and is exploring whether a national approval pathway is feasible, rather than navigating a patchwork of state-by-state regulatory rules.

A central unanswered question is whether the Food and Drug Administration (FDA) will step in.

Traditionally, states regulate the practice of medicine while the FDA oversees medical devices. But AI systems that make clinical decisions blur that boundary.

Lowell Schiller, a former chief counsel for the FDA, said that because Doctronic’s AI is designed to renew patient prescriptions, it is effectively practicing medicine and would typically fall under state regulation. At the same time, the FDA has said it believes it has authority to regulate AI as a medical device if it is used to diagnose, treat, or prevent disease.

Schiller added that even if the FDA has such authority, it could choose to hold off on taking action, citing medical marijuana as an example of where the agency has deferred to state laws rather than enforcing federal regulation.

If the FDA determines that Doctronic’s AI is being marketed without appropriate authorization, the agency could seek to bring the technology into compliance. Under President Donald Trump, the FDA has shown an increased willingness to assert oversight in emerging health technologies.

Earlier this year, the agency sent a warning letter to health-wearable company Whoop, stating that it could not market its blood-pressure estimation technology without FDA approval. For low- to moderate-risk medical devices without existing predicates, the FDA generally aims to issue a marketing authorization decision within 150 days, though the process can take longer.

Zach Boyd, director of Utah’s artificial intelligence policy office, said that historically the regulatory boundary has been clear: states regulate the practice of medicine, while the FDA regulates medical devices.

“Now we’re in this weird place where there are devices—maybe you could call them devices—that are purporting to practice medicine,” Boyd said. He added that Utah’s approach has been to focus on its own regulatory authority, while leaving it to the FDA to determine how it will address AI systems that blur those traditional lines.

The FDA declined to comment on the program, stating it falls outside the agency’s regulatory purview—for now.

Q: Is AI fully replacing doctors in this program?
A: No. The AI handles routine renewals for approved medications, but escalates cases to physicians when uncertainty arises, and human doctors review early prescriptions for validation.

Q: Why hasn’t the FDA intervened yet?
A: The regulatory line between medical devices and the practice of medicine remains unclear for AI systems, leaving oversight largely with states at this stage.

Q: What medications are excluded?
A: Pain management drugs, ADHD medications, injectables, and other higher-risk prescriptions are excluded for safety reasons.

Q: Who is legally responsible if the AI makes a mistake?
A: Doctronic has secured a malpractice insurance policy covering its AI system, treating it as accountable at the same level as a physician. The company has stated it assumes responsibility for patient outcomes regardless of whether an issue is caused by AI or human oversight.

Q: Could this expand beyond Utah?
A: Doctronic is in discussions with several other states, including Texas, Arizona, and Missouri, and is evaluating whether a national approval pathway is possible. Any broader rollout would likely depend on how state regulators and the FDA ultimately define oversight for AI-driven medical decision-making.

Utah’s experiment places AI directly into one of the most trust-based relationships in healthcare: prescribing medication. If the pilot proves safe and effective, it could redefine how routine care is delivered—particularly for patients managing chronic conditions who face long wait times, limited access to providers, or frequent gaps in medication.

The implications extend beyond efficiency. In areas experiencing physician shortages or limited access to care, AI-driven prescription renewals could reduce delays that have real health consequences, from unmanaged symptoms to avoidable hospital visits. For patients who rely on consistent medication, timely renewals can be the difference between stability and crisis.

At the same time, the program highlights unresolved questions about accountability, oversight, and where responsibility lies when clinical decisions are automated. How regulators address those questions—at both the state and federal level—will shape whether AI prescribing remains a narrow experiment or becomes a broader tool for expanding access to care.

For patients, providers, and policymakers alike, Utah’s experiment is not just about technology replacing tasks. It is a test of whether automation can responsibly support human healthcare systems at a moment when demand is rising and access remains uneven.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.