Legion Health AI

The Big Picture: Utah is home to the first-in-the-nation Office of Artificial Intelligence Policy, which oversees our AI Learning Laboratory program.
Regulatory Mitigation, Not Endorsement: The goal of this program is to facilitate the safe deployment of novel AI applications by removing regulatory barriers temporarily, while closely monitoring the results to safeguard public safety. Entering into a regulatory mitigation agreement with a participant does not constitute an endorsement or approval of their specific technology by the State of Utah.
Evidence-Based Policy: By testing these technologies in a controlled, 12-month environment, we can gather qualitative and quantitative data on AI's real-world impact before making permanent changes to state law.

The Access Crisis: Most counties in Utah have designated mental health provider shortages, leaving up to 500,000 residents without adequate access to behavioral healthcare.
Reducing Bottlenecks: Routine prescription renewals are high-volume and high-friction. By safely automating the renewal process for maintenance medications, we are allowing patients to get the care they need much more quickly and affordably, which ultimately prevents avoidable mental health deterioration.
Freeing Up Human Clinicians: Using AI for routine administrative tasks and straight renewals allows licensed healthcare providers to focus their time on more complex, higher-risk patient needs, expanding overall access to care in our state.

General vs. Specialized Care: Our previous agreement with Doctronic focused broadly on primary care capacity, utilizing an AI system to process renewals for a wide formulary of nearly 200 common medications, spanning from asthma to hypertension. The Legion agreement represents a narrowly tailored, specialized approach: it strictly targets the state's mental health shortage by focusing exclusively on non-controlled, maintenance psychiatric medications, such as SSRIs for depression and anxiety.
Consistent Safety Standards: Both programs share a strict prohibition on the renewal of controlled substances and utilize a phased human-review process to rigorously benchmark the AI against licensed human providers.

Term & Liability: The mitigation is granted for a 12-month period. Crucially, while the state waives specific Division of Professional Licensing (DOPL) enforcement actions regarding the use of AI to authorize telehealth renewals, the agreement does not waive any liability or legal remedies if a patient is harmed.
Strict Guardrails & Exclusions: The AI is only authorized to handle renewals of previously prescribed medications. It cannot issue new prescriptions, change doses, or prescribe controlled substances, benzodiazepines, or antipsychotics.
Hard Stops for Human Escalation: The AI employs conservative eligibility gates. It will immediately escalate the patient to a human clinician if it detects suicidality, severe adverse effects, indications of mania, pregnancy, or if the patient simply requests a human review.
Phased Auditing Requirements: Legion must prove the AI's safety in rigorous stages. For the first 250 requests, a licensed physician must review the case before the prescription is sent to the pharmacy, requiring a >98% agreement rate. The next 1,000 requests undergo intensive retrospective review, requiring a >99% agreement rate, before moving to ongoing monthly randomized sampling.
Monthly Reporting: Legion must submit detailed monthly reports to the Office of AI Policy, outlining the number of accepted/denied renewals, concordance rates with human physicians, and any user complaints or adverse health outcomes.