Doctronic wins Utah nod to renew 190 prescriptions via AI

A US state has quietly handed a chatbot something no state has handed a piece of software before, the legal authority to renew a real prescription. The AP reports that Utah's program, run through a startup called Doctronic, launched earlier this year and now covers a list of 190 medications, including blood thinners.

The mechanic is straightforward. The chatbot confirms the patient's identity, asks about their prescriptions and medical history, and cross-checks a national pharmacy database. If nothing looks off, it sends the refill to a local pharmacy. If anything looks off, it hands the conversation to a telehealth doctor who works for Doctronic. Adam Oskowitz, a Doctronic co-founder, framed the pitch as meeting patients where they need healthcare.

Doctors are not sold. Dr. Eric Bressman at the University of Pennsylvania told reporters, 'We have crossed a threshold in terms of giving something that is not human a medical license.' Dr. Alan Smith, who chairs Utah's Medical Licensing Board, said the 11-member board was essentially informed the pilot was going ahead without its input. The American Medical Association said renewals 'aren't routine checkboxes.' Utah has already pulled medications from the list, reportedly including a drug for irregular heartbeats, after safety pushback.

The reason this matters beyond one state is that the federal backstop is not really there. An FDA spokesperson said the agency has not authorized any AI chatbots but is 'committed to encouraging medical innovation,' which in practice is a hands-off posture. Utah's regulatory sandbox waives the rules that would otherwise restrict prescribing to licensed professionals, which means the operative rulebook for AI-driven prescribing right now is a state-level pilot, not a federal standard. Daniel Aaron at the University of Utah's law school warned that expanding beyond current evidence risked, in his words, compromising public trust and fueling backlash, and that is the regulatory risk to watch.

The honest caveat is what the reporting does not quite give you, the actual escalation rate to human doctors in production, what adverse-event data Utah is collecting, and where malpractice liability sits when the AI approves a refill and something later goes wrong. What the reporting does flag, worth holding onto, is that human doctors currently review all refill orders during the pilot phase, and that Doctronic's own study put its diagnostic accuracy on par with physicians 80% of the time, a company-authored number. The upside if this works is real for chronic-care patients in Utah and for the next state watching. The downside if it doesn't is a template other regulators quickly borrow from.