Utah pilot lets Doctronic AI renew some chronic medication refills

In short: Utah has started a one-year pilot program that allows Doctronic’s AI chatbot to renew refills for a limited list of long-term medicines, under special state rules.

What happened

Utah launched the pilot in January 2026 in partnership with Doctronic. The AI system can renew existing prescriptions for about 190 to 192 medications used for chronic conditions, such as blood pressure drugs, diabetes medicine, birth control, and some antidepressants like SSRIs.

The program does not allow the AI to write brand new prescriptions. It only renews prescriptions that were originally prescribed by a licensed clinician. It also excludes controlled substances and several higher risk categories, including painkillers, ADHD medications, injectables, and medicines meant for sudden, short-term problems.

Patients request a 30, 60, or 90-day refill through a chatbot and pay $4. The chatbot checks identity and Utah residency, then asks about symptoms, side effects, whether the patient is taking the medicine as directed, and possible warning signs. If the case looks risky, it is sent to a human clinician instead of being renewed automatically.

Utah is running this through a “regulatory sandbox,” which is a supervised test environment where the state temporarily relaxes certain rules so it can study what happens (like letting a new kind of service operate in a controlled trial). The rollout has phases, including early physician sign-off on the first set of renewals in each drug group, then later reviews and random safety checks. Doctronic says it carries malpractice insurance similar to physician standards.

Why it matters

For patients, this could make routine refills faster and cheaper, especially for people who need the same medicine month after month. For the state, it is a real-world test of how much medical decision-making can be safely handled by software, and where humans still need to be involved. Critics, including some doctors and pharmacists, argue the AI’s decisions may be hard to evaluate from the outside and could create safety and accountability questions.

Source: The Verge AI