# med.org.ai

> 92.3 million Americans live where there are not enough primary-care physicians
> to go around. We are writing down what a physician's supervised AI may now do
> about that — and what stays with the licensed human.

med.org.ai is a program of the Org.AI Foundation (foundation.org.ai). It is the
medical face of the Foundation's map: which acts of medical care are reserved to
licensed humans, which acts a physician may supervise a machine in doing, and
what a regulator — and a court — would need to see before recognizing the
physician's supervised agent inside the confidential encounter.

## Status

Nothing on this site is a medical service. No patient is served here. This page
is the map, published first, so that the people who regulate the practice of
medicine can read it before anyone asks them for anything. No sandbox
application has been submitted, and no rule change has been proposed anywhere.

## The three claims

1. ALREADY IN THE ROOM. The confidential encounter has never been just two
   people. Medicine admits helpers under the physician's direction: medical
   students (Medicare has billed on student documentation, with physician
   verification, since 2018), physician assistants on delegated authority,
   human scribes — and, since 2023, ambient AI scribes that listened and
   drafted the note in millions of visits under ordinary HIPAA
   business-associate agreements, roughly 2.5 million encounters across 7,260
   physicians in one medical group alone (the Permanente Medical Group,
   reported in NEJM Catalyst). Every rung is a non-physician admitted on the
   same conditions: the physician directs, reviews, and answers for it. The
   missing rung is the agent that acts — prepares the renewal, runs the
   intake, makes the follow-up call — not the agent that transcribes.

2. THE PHYSICIAN STAYS. The order of events is the design: a licensed
   physician first, then the machine — never the reverse. The agent is the
   physician's supervised instrument in medical care — never a caregiver,
   never a therapist. The physician is named and carries personal professional
   responsibility, which is where the Federation of State Medical Boards' 2024
   policy puts accountability for AI in clinical practice. The patient is told
   and accepts. Diagnosis, the prescribing decision, and escalation stay with
   the physician. Every direction is logged — which physician directed which
   task for which patient, when — in a tamper-evident audit trail held as
   medical-record material. No training on patient data, no sale of it: an
   architecture that can be inspected, not a policy promise.

3. THE CONFIDENTIALITY QUESTION. Federal law contains no general
   physician-patient privilege — the Supreme Court has recognized only the
   psychotherapist-patient privilege (Jaffee v. Redmond, 1996) — so the
   evidentiary shield lives state by state. HIPAA already permits disclosures
   for treatment and already admits technology vendors as business associates
   under contract. And Utah Rule of Evidence 506, with counterparts in Texas,
   New Mexico, and North Dakota, already extends the confidence to persons
   "participating in the diagnosis or treatment under the direction of the
   physician." No state has yet said whether those words reach a supervised,
   patient-consented AI agent that acts in care. We think the answer runs
   through supervision, not around it.

## Utah — the first venue (med.org.ai/ut)

Under its 2024 Artificial Intelligence Policy Act (SB 149), Utah's Office of
Artificial Intelligence Policy signs regulatory mitigation agreements —
published contracts admitting a specific AI deployment into regulated practice
under specific conditions. It has twice admitted supervised, agentic clinical
AI: a January 2026 agreement authorizing AI-prepared prescription renewals for
roughly two hundred non-controlled medications, co-signed by the state's
physician licensing division, with physician oversight phased from full
pre-issuance review toward audited samples; and an agreement under which an AI
prepares psychiatric-medication renewals with patient consent, audits, and
escalation triggers. The office that signs the agreements and the division that
licenses physicians sit in the same Department of Commerce.

The ask being prepared has three instruments: a regulatory mitigation agreement
for a physician-supervised care agent with published metrics and a written
conversion trigger; delegation guidance from the Division of Occupational and
Professional Licensing recognizing a certified, physician-supervised,
patient-accepted agent as a permissible delegate under the practice act; and a
committee note or one-clause amendment confirming that an authorized,
physician-directed agent falls within Rule 506's existing clause. None of these
instruments is the Foundation's to file: physicians and their institutions
petition; the Foundation prepares the research, the drafting, and the audit
infrastructure, and assists.

## Sources

Every claim above names a primary source: commerce.utah.gov (the Office of
Artificial Intelligence Policy's published agreements and pilot pages,
including the January 2026 renewal agreement in full), the Utah state courts'
published rules (Utah Rule of Evidence 506), le.utah.gov (SB 149, 2024; HB
452, 2025), data.hrsa.gov (shortage-area designations: 92.3 million people in
primary-care HPSAs, 137.1 million in mental-health HPSAs), catalyst.nejm.org
(the Permanente Medical Group ambient-AI figures), fsmb.org (the April 2024
policy on AI in clinical practice), ama-assn.org (physician AI-adoption
surveys), ecfr.gov (HIPAA's treatment-disclosure and business-associate
provisions), and the published opinion in Jaffee v. Redmond.

## Licence and use

Nothing on med.org.ai is generated at request time. This page and this file are
static, self-contained, and make zero external requests. Nothing here is
medical advice, and nothing here creates a physician-patient relationship.