Nolla Health launches AI‑driven acne prescription pilot in Utah

On 5 October 2026, Nolla Health disclosed a pilot program in Utah that uses an artificial‑intelligence engine to generate first‑line prescription recommendations for adults suffering from mild to moderate acne. The service is offered at a flat rate of $4.99 per month and requires users to upload facial photographs from five angles together with a health questionnaire, creating a comprehensive visual and clinical profile for each participant.

The AI evaluates the images and questionnaire responses, then selects a treatment plan from a pre‑approved pool of eight medications. According to the company, the algorithm produces a recommendation within ten to fifteen minutes, after which a human clinician reviews the output before it is delivered to the patient, ensuring that the automated suggestion is vetted by professional judgment.

Three‑phase oversight model

The pilot is structured into three distinct phases, each with increasing autonomy for the AI and decreasing frequency of human oversight. In the first phase, covering the first one hundred cases, two physicians must examine each recommendation before it is sent to the patient, and this review must continue for at least four weeks, providing a rigorous safety net during the initial rollout.

The second phase expands the cohort to five hundred cases. Here, a medical review occurs weekly after the prescription is issued and must be sustained for a minimum of eight weeks. This weekly check is intended to capture any adverse reactions or treatment failures early in the process, allowing rapid corrective action if needed.

The third phase, which the company describes as the "steady‑state" stage, retains a monthly audit of at least ten percent of all cases and any alerts generated by the system. This ongoing monitoring is designed to ensure long‑term safety and compliance with clinical standards while granting the AI greater operational latitude.

Transition criteria and regulatory safeguards

Advancement from one phase to the next is contingent upon meeting three conditions outlined in Nolla's protocol: (1) at least 95 percent agreement between the AI’s recommendations and the reviewing physicians, (2) no occurrence of serious adverse events, and (3) explicit approval from Utah state health authorities, thereby embedding state oversight into the scaling pathway.

The AI is programmed to avoid unrestricted prescribing. It can only choose from the eight pre‑approved treatments and must flag cases that involve pregnancy, severe acne, or any clinical uncertainty for immediate escalation to a qualified dermatologist, ensuring that high‑risk scenarios are handled by experts.

Company‑reported performance metrics

Nolla Health claims a 96 percent concordance rate between its AI suggestions and the clinicians who reviewed them during internal testing. The company also cites data from eleven Utah counties that lack a resident dermatologist, noting an average wait time of 61 days for specialist appointments in those areas, underscoring the access gap the pilot seeks to address.

• $4.99 monthly subscription
• Five‑angle facial photo upload
• Health questionnaire
• AI selects from eight approved acne treatments
• Physician review required for first 100 cases

The pilot does not replace a dermatologist; instead, it aims to fill a care gap in underserved regions by delivering evidence‑based, low‑cost prescriptions quickly while maintaining a safety net of professional oversight, thereby complementing existing dermatology services rather than supplanting them.

If the program meets its transition thresholds, Nolla plans to seek broader adoption across other states with similar dermatology shortages. The company has indicated that scaling would involve adapting the protocol to each state’s regulatory framework and potentially expanding the treatment catalogue, while preserving the core oversight structure.

For English‑speaking healthcare organizations, the pilot demonstrates a concrete model for integrating AI into routine dermatologic care without relinquishing clinical responsibility. It offers a template for how AI can accelerate prescribing workflows, reduce patient wait times, and lower costs, provided that robust review mechanisms and state approvals are firmly in place.

The initiative also highlights the importance of data privacy and consent, as participants must grant permission for their facial images and health information to be processed by the AI engine, with all data stored in compliance with HIPAA and Utah state regulations.

Stakeholders have praised the transparency of the three‑phase model, noting that the staged reduction in human oversight is tied to measurable safety outcomes rather than arbitrary timelines, which could set a precedent for future AI‑driven medical pilots.

Critics, however, caution that reliance on a limited medication pool may not address all nuances of acne pathology, and they call for ongoing research to validate long‑term efficacy across diverse patient populations.

Overall, the Utah pilot stands as a pioneering effort to blend artificial intelligence with traditional medical practice, aiming to deliver timely, affordable acne treatment while rigorously safeguarding patient health through layered clinical review.

From Salt Lake City to the remote mountain counties, the program reflects Utah’s commitment to leveraging technology to bridge healthcare deserts, offering residents a faster, cost‑effective pathway to manage a common skin condition that often goes untreated due to specialist scarcity.