
AI prescribing gets a state-supervised test in Utah
Key Takeaways
- Utah authorized an AI to issue initial acne prescriptions under physician oversight, establishing a state-supervised framework that includes medical board engagement and multi-stakeholder external review.
- A tiered governance model reduces prospective physician sign-off as volume grows, shifting from per-prescription approval to retrospective daily review and then weekly sampling.
Nolla Health says it is the first company cleared for AI to issue initial prescriptions. Physician review shifts from every prescription to weekly spot checks as the pilot grows.
Utah has become a testing ground for a question developing in healthcare artificial intelligence (AI): whether software can make a prescribing decision on its own.
Nolla Health announced Oct. 5 that it has become the first organization in the country to receive regulatory approval for AI to issue initial prescriptions. Its
The company draws a line between its system and telehealth platforms that use AI only to collect information before handing patients off to clinicians. Nolla's AI is designed to independently issue initial prescriptions within tightly defined, physician-approved treatment protocols. The company describes the pilot as a new model for delivering routine healthcare safely, quickly and affordably at scale.
"AI will matter most in healthcare when it helps people get care they otherwise would not receive," Luis Wenus, co-founder and CEO of Nolla Health, said in the news release. "Growing up with a doctor in the family, I saw how powerful it was to have trusted medical help within reach, and millions of people do not have that today."
A state-supervised path to approval
Nolla Health said it worked for months with the Utah Office of Artificial Intelligence to secure approval for an AI system authorized to issue initial prescriptions under physician oversight. The company also met and worked with the Utah Medical Board before launch so the board's feedback would shape the final program.
Before launch, the pilot went through multiple rounds of external review by practicing physicians, dermatologists, key stakeholders and members of Utah government, according to the release.
"The Nolla Health pilot program represents a significant step forward in leveraging AI to improve healthcare access for Utahns, while maintaining rigorous oversight and safety," Office of Artificial Intelligence Policy Director Zach Boyd said in the news release. "Our office facilitates innovative AI solutions while ensuring that public health and patient safety remain paramount."
Oversight that steps down as volume grows
The program's governance model ties the level of physician review to the number of patients treated. According to Nolla Health:
- Stage one covers the first 100 patients. A licensed physician reviews and approves every AI-generated prescription before it reaches the patient.
- Stage two runs through the first 500 patients. The AI issues prescriptions autonomously, and a physician conducts a daily review after the fact.
- Stage three follows. A physician reviews a sample of prescriptions once a week.
Zaid Fadul, M.D., chief medical adviser for Nolla Health and part of the physician team overseeing the pilot, framed the arrangement as a question of control.
"The real test for AI in healthcare isn't how smart it is. It's whether someone other than the company can shut it off," Fadul said in the news release. "In Utah, the state can, and physicians can — that's the standard every AI in medicine should meet."
Fadul's comments touch on
Guardrails and performance claims
The system cannot generate open-ended prescriptions or improvise treatments, according to the company. It selects from pre-approved treatment pathways using structured clinical data and objective analysis of a facial scan. If the AI cannot confidently select a treatment, the patient is directed to a licensed physician.
"People are right to worry about AI in medicine that nobody can control. This program was built to be the opposite," Fadul said. "The AI can't improvise. It can only choose from a short list of physician-approved topical treatments. The moment a patient falls outside those limits, it stops and hands them to a doctor."
Humans also handle patient questions. Patients can message a licensed physician through the app at any time and never speak with AI when they reach out for support, the company said.
Nolla Health said licensed clinicians agree with the AI's treatment recommendations in more than 96% of real-world cases. The remaining cases reflected minor adjustments, such as changes in the strength of a topical medication, rather than fundamental disagreement with the treatment, according to the company.
The access and workforce case
The company built its case for the pilot on gaps in dermatology access. In Utah, an estimated 100,000 people live with active acne at any given time, but only 10% ever see a dermatologist, according to the release. The average wait in Utah is 61 days, nearly double the national median, and 11 counties have no dermatologists. Visits can cost up to $300.
Nationally, more than 80% of U.S. counties are considered healthcare deserts, affecting more than 120 million people, or about one in three Americans, the company said.
Nolla Health positions the pilot as
The company chose acne as a starting point because it is a well-understood, low-risk condition with decades of clinical evidence behind it, according to the release.
"That's the future Nolla is building towards: AI helping high-quality healthcare reach far more people," Wenus said.
Funding and footprint
Nolla Health, headquartered in New York, now operates in 44 states, and the Nolla Derm app has been downloaded by more than 175,000 people, according to the company.
The company expanded to the United States in September 2025 with a $4.5 million pre-seed funding round led by General Catalyst, with participation from SNÖ Ventures and Commure founder Diede van Lamoen, according to a company blog post at the time. In that post, Nolla said it had treated more than 50,000 patients in Norway in six months while reducing clinician time per patient tenfold.
"At General Catalyst, we believe AI-native healthcare models can make care more accessible and affordable for patients everywhere," Neeraj Arora, managing director at General Catalyst, said in the 2025 post.
At that time, the company said it planned to use the funding to grow its clinical and engineering teams and develop additional condition-specific apps.
Related to this article








