When AI Writes the Prescription: Utah's Autonomous AI Pilot and the Regulatory Challenges Ahead
Utah is piloting autonomous AI systems capable of making clinical decisions—including writing prescriptions—without direct physician oversight. The initiative is drawing scrutiny from healthcare lawyers who say current regulatory frameworks lag far behind the technology.
A central concern is medical licensure. State medical boards typically license individual practitioners, not software. When an AI system independently prescribes medication, it remains unclear which jurisdictional rules govern its "practice" and whether a single state license could apply to an algorithm operating across state lines.
Liability is equally unresolved. If an AI-generated prescription causes harm, questions arise over whether liability falls on the healthcare provider, the hospital system deploying the software, the AI developer, or the patient’s insurer. Traditional malpractice doctrine assumes a human practitioner; apportioning fault among multiple corporate and technical actors complicates established legal precedent.
The corporate practice of medicine doctrine—which restricts non-physician entities from exercising clinical judgment—poses another barrier. When an AI makes a treatment decision, is it acting as a licensed physician proxy, or is the corporate entity effectively practicing medicine through software? Courts and regulators have not clearly answered this question.
Healthcare legal analysts note that Utah’s pilot may ultimately require legislative fixes, updated board rules, or newSafe Harbor provisions for AI-assisted care. Without clear guidance, providers may be reluctant to deploy autonomous systems despite potential efficiency gains, and patients may lack recourse if something goes wrong.