Utah's Office of Artificial Intelligence Policy approved a staged-autonomy model for AI decision-making — intensive human review that tapers to spot-checks o…
0:00 On October 5th, 2026, Utah became the first US state to authorize an artificial intelligence system to issue initial prescriptions without a doctor reviewing every case in real time. The system is limited to topical acne medications for mild to moderate cases. [music] This is not a refill program. It is the first time AI has been allowed to write a brand new prescription under a state regulatory agreement. The app is called NolaDerm, made by Nola Health. Utah residents 18 and older can download it, verify their identity with a government ID and selfie, complete a short intake questionnaire, and take a five-angle face scan. According to the company, the process takes about 10 to 15 minutes. If the AI determines the patient has mild to moderate acne and meets the criteria, it can generate a prescription from a fixed list of topical treatments only, including tretinoin, adapalene, clindamycin with benzoyl peroxide, benzoyl peroxide, and azelaic acid.
1:00 Oral medications, isotretinoin, and severe acne are excluded. The pilot price is $4.99 per month. There is staged physician oversight. [music] In the first stage, covering the first 100 patients, two licensed Utah physicians must independently review and approve every AI-generated prescription before it reaches the patient. In later stages, the level of real-time review decreases, moving first to post-issuance review, and eventually to sampling a percentage of cases. Nola Health states that licensed clinicians have agreed with the AI's treatment recommendations in approximately 96% of recent real-world cases. Most disagreements were minor adjustments in strength rather than changes to the treatment itself.
1:50 Utah has significant access gaps in dermatology. The company and state materials cite roughly 100,000 people in Utah with active acne at any given time with only about 10% ever seeing a dermatologist. The average wait for an appointment is reported at 61 days, nearly double the national median. 11 counties have no dermatologist at all. In-person visits can cost up to $300. More than 80% of US counties are considered healthcare deserts of some kind affecting over 120 million people. This pilot is part of Utah's AI sandbox program which allows temporary regulatory relief for controlled testing of AI systems under state oversight. Earlier in 2026, Utah authorized a separate AI system, Doctronic, [music] to handle certain prescription refills.
2:44 That program drew public criticism from members of the Utah Medical Licensing Board who raised concerns about patient safety and the level of clinical oversight. The Nola pilot goes further by allowing initial prescriptions, not only renewals. The agreement includes monthly reporting requirements, adverse event reporting within 24 hours, and third-party [music] auditors. Liability and accountability questions remain under discussion among physicians and regulators. Nola Health previously raised $4.5 million in funding including from General Catalyst. Its app had already been downloaded by more than 175,000 44 states under a model that still required clinician approval. The Utah pilot is the first time the company has been authorized to let the AI issue the initial prescription under the staged oversight described earlier.
3:39 The core factual questions this raises are straightforward. Does the [music] staged review model provide sufficient safety for low-risk topical acne treatment? Will the reported 96% agreement rate hold as the volume increases and human review decreases, and how will other states respond if the pilot data shows acceptable outcomes? Those are the documented facts as of today. What is your view on allowing AI to issue initial prescriptions under this kind of staged oversight? Leave your thoughts in the comments.