Nolla Health starts issuing AI-generated acne prescriptions in Utah

Healthcare startup Nolla Health has started issuing AI-generated prescriptions for acne to patients in Utah, after state regulators signed off on the pilot. The company says it is the first in the country to let an AI system write an initial prescription, rather than just renew one a doctor already approved, according to a Bloomberg report cited by The Verge.

The process runs through Nolla Health’s app. A user scans their face with a smartphone camera, and the company’s AI system analyzes the severity of their acne before generating a prescription. Physician oversight is built into the pilot but loosens as it scales. For the first 100 patients, two physicians review and approve each AI-generated prescription before it goes out. That shifts to after-the-fact review for the next 500 patients. Past that point, Nolla Health says physicians will check a sample of at least 10 percent of prescriptions every month, plus every case that involves an escalation or a side effect.

Utah opened the door to this kind of system earlier this year, when it began allowing AI to renew certain prescriptions. Several other healthcare companies are already testing AI-assisted renewals in the state. Nolla Health’s pilot goes further by having its AI handle the first prescription a patient receives, not just a repeat order, which the company says makes it the first in the country to do so.

The service costs $4.99 a month and is limited to Utah residents who are at least 18 years old and have mild-to-moderate acne. Nolla Health says its AI can currently prescribe eight different skin treatments. If the system cannot confidently choose a treatment for a given case, the company says it will route that patient to a physician instead of guessing.

In its press release, Nolla Health frames the pilot as a complement to dermatologists rather than a substitute for them, saying that handling routine, low-risk cases through AI frees physicians to focus on complex, high-risk patients. Neither the company’s announcement nor the Bloomberg report describes how the AI model was built, what data trained it, or how its accuracy compares with a dermatologist’s judgment, so those details remain unknown for now.

The pilot’s own numbers mark the points worth watching: the shift from upfront physician sign-off to after-the-fact review once the patient count passes 100, and the further shift to a 10 percent monthly sample once it passes 500. Those thresholds, not a launch date, are what will show how much oversight Nolla Health keeps in place as the program grows.