Utah gives startup Nolla Health permission to start issuing automated AI prescriptions for acne treatments

Utah has become the first state in the U.S. to grant an artificial intelligence startup permission to automate the approval of prescription refills, with minimal oversight from human doctors.

In a blog post today, Nolla Health said it has received approval from the state to launch a pilot project that will allow acne patients to talk to an AI agent, upload a photo of their face, and then automatically receive a prescription without having to speak to a doctor. “This makes Nolla the first end-to-end AI doctor ever,” said Luis Wenus, founder of Nolla Health, in a post on X today.

Through the Nolla Derm app, available on iOS devices, an AI model can select which drug to prescribe to each user from a shortlist of doctor-approved topical treatments, the company said. All requests from patients that fall outside of these strict limits are referred to a real, licensed doctor.

When users first download the app, they must go through an intake process before they can actually speak with the AI ​​doctor. Once approved, they can tell the app their acne condition, upload a facial scan showing how bad their acne really is, and the AI ​​will then decide on an appropriate treatment plan.

The prescription can then be picked up at your local pharmacy or sent directly to your home. The app also tracks the patient’s progress through a daily skin scan and updates their treatment plans monthly depending on how well the app is working.

“Previously, a licensed physician signed every single prescription before it was dispensed,” Wenus explained in a video introducing the app.” Our AI can now make this prescribing decision itself.”

Today, Nolla Health became the first organization in the US (and potentially the world) to receive regulatory approval for an AI to write initial prescriptions.

This makes Nolla the first truly integrated AI doctor ever. pic.twitter.com/xn6rvb8IyW

– Luis Wenus (@luiswenus) October 5, 2026

Preliminary approval

The service is being made available as a pilot after being approved by the Utah Office of Artificial Intelligence Policy, the Utah Division of Professional Licensing, the Utah Medical Association and the American Academy of Dermatology. Under the agreement, the company’s AI system is only allowed to issue prescriptions for topical acne treatment creams. She is prohibited from prescribing isotretinoin and oral medications.

The pilot program consists of three phases. In the first phase, two doctors approve each AI-generated prescription before it is sent to the user. Once this has been done on at least 100 patients and there are no serious problems, it moves to the second phase, where the AI ​​writes the prescriptions directly without human consent and the doctors review them after the fact. This phase will include 500 patients, Nolla Health said.

Finally, in the third phase, human doctors simply review at least 10% of all Nolla’s automated prescriptions each month. The pilot is being launched as part of Utah’s AI Learning Lab sandbox program, which exempts AI companies from state laws to test new products under supervision. “Participation in our AI Sandbox does not imply that the State of Utah approves, endorses, certifies, or vouches for the product,” the Utah Office of Artificial Intelligence Policy Authorized Pilots page states.

Acne as proof of concept

In an interview with Inc.com, Nolla Health chief medical advisor Zaid Fadul said that acne treatment is just the starting point. The startup has much bigger ambitions for the future, he said. If the pilot is successful, the aim will then be to introduce automated prescribing services for other common conditions that are well understood by doctors, can be easily assessed remotely and can be treated with medications with a strong safety profile.

“Beyond dermatology, I think conditions like the flu and uncomplicated urinary tract infections are worth investigating, especially because testing can provide objective information that can be considered along with a patient’s symptoms and medical history,” Fadul said. “The key is to have very clear guardrails.”

In 10 years, Fadul said, he expects AI to be able to manage a large portion of humans’ routine health checks, including diagnoses, treatments, prescription refills and routine tests. However, he does not expect AI to be used for more serious illnesses and life-changing diagnoses, or for major procedures and high-risk medications.

Some medications, such as stimulants, opioids and benzodiazepines, are too risky for AI to prescribe, he said, because there is a risk that some people might try to manipulate the models that do the assessments. “Frankly, it may be a long time before we have enough evidence and adequate safeguards to support independent prescribing of some of these drugs,” Fadul said.

Image: Nolla Health