AI Chiefs Enter the Hospital C-Suite: Florida 2026

A few years ago, the most powerful person deciding how you would practice was someone you had met: a department chair, a CMO who had once rounded on the same wards you did. Walk the executive floor of a large health system today and you will find a new office, and the name on the door is one most physicians cannot match to a face. The title reads Chief AI Officer, or some close cousin of it. The person inside is making decisions that will reach your exam room long before you are asked for an opinion.

This is not a rumor about the future. The org chart has already been redrawn. The question for Florida physicians is whether we walk into the room these people now occupy, or wait to be handed the result of meetings we were never invited to.

The New Names on the Org Chart

The American Hospital Association in February 2026 described health systems “rewriting the org chart” to elevate technology leadership into the C-suite. The titles vary, and the variation itself tells you the field is still forming. You will see Chief AI Officer, Chief Health AI Officer, Chief Data and Artificial Intelligence Officer, Chief Digital Transformation Officer, and Chief Transformation Officer. Underneath the different words sits the same intent: give artificial intelligence, data, and automation a single owner who reports to the CEO and answers to the board.

The examples are no longer confined to coastal academic centers. Cleveland Clinic named its first Chief AI Officer in 2024 to own enterprise AI strategy, governance, and safety. Cedars-Sinai and Hackensack Meridian followed with their own. NYC Health and Hospitals, the largest public system in the country, created a Vice President and Chief Data and Artificial Intelligence Officer to run data and real-time intelligence across a vast safety-net network. Mass General Brigham spread the work across a CIO, a chief medical information officer, and a chief data science officer rather than crown one person. And the trend reached rural America in January 2026, when Sanford Health, a 56-hospital system serving the Upper Midwest, installed Tommy Ibrahim, M.D., as its first Executive Vice President and Chief Transformation Officer, charged with using AI to make value-based care work across long distances and thin margins.

Notice who some of these people are. Several carry an M.D. The most consequential version of this role is not a technologist who learned a little medicine. It is a physician who learned to run the technology, and who now sits where the CMO used to sit alone.

What Their Decisions Mean at the Bedside

When a Chief AI Officer chooses an ambient documentation vendor, sets the threshold at which a sepsis model fires, or decides which patient populations a risk algorithm scores first, that person is writing clinical policy. They may not call it that. It is.

Consider what flows downstream from a single procurement choice. The American Medical Association’s augmented intelligence research found that physician use of AI jumped to 66% in 2024 from 38% the year before, a 78% increase in twelve months, with most of that growth in clinical documentation. The tools are already in our hands. The unsettling part is who configured them. That same body of AMA survey work found that only about 8% of physicians said their organization’s AI decision-making was clear and that they understood the governing policies. We are using systems we did not design, tuned to metrics we did not pick.

The optimization target is everything. An ambient scribe chosen to relieve a physician’s after-hours charting is a different tool than the same scribe chosen to push three more patients onto the afternoon schedule. The software is identical. The intent embedded by the executive who bought it is not. When administrators and payers select the metric, throughput tends to win, the same incentive that fueled the public backlash against UnitedHealth we covered this spring. When physicians hold a real vote, the metric bends back toward the patient. This is the seam where the practice of hospital medicine is quietly being reshaped, and most of us are not standing in it. It is the same dynamic that drove the AI scribe consent litigation we covered earlier this year, when vendors moved faster than anyone had thought to ask whether the tool was legal in Florida.

In Florida, our aging population leans harder on the systems most eager to automate, and our physician distribution problems mean rural and underserved counties will get whatever the algorithm decides to prioritize. A risk model that under-scores a county with sparse historical data does not announce itself. It just sends the resources somewhere else.

How to Make Yourself Hard to Sideline

The reflex when a new power center appears is to resist it or ignore it. This is not the time to make a misstep. The physicians who keep their standing will be the ones who ask sharper questions than the people buying the software, and who make those questions impossible to route around. Start here.

1. Ask what every new tool is optimized for, and ask in writing. Before your hospital deploys an AI system, send the medical staff office one question: what outcome was this tool selected to improve, and who measures it? If the answer is throughput, you have found the fight worth having. If no one can answer, you have found a governance gap you can fill.

2. Get a physician onto the AI governance committee. Most systems now run an AI oversight committee. The AMA argues that a workable committee needs clinical, technical, and operational voices on it at minimum. Do not assume a doctor is in the room. Find out who occupies this seat, and if the seat is empty or ceremonial, volunteer or nominate a colleague who will actually get engaged.

3. Audit one model you already rely on. Pick the sepsis alert, the readmission score, or the scribe you use daily, and ask for its validation data on your patient population, its false-positive rate, and the date it was last revalidated. The act of asking changes who gets consulted next time.

4. Tie your demands to organized medicine. A single physician asking is a complaint, but a medical staff backed by the Florida Medical Association is a negotiation. The FMA can press for clinician representation in procurement at the system and state level in a way no individual can.

5. Make yourself the translator, and do not do it alone. The physician who can sit between the Chief AI Officer and the medical staff, fluent in both, becomes the person neither side can proceed without. That is the kind of standing you build rather than wait to be granted. It is also the work The Atlas Accord exists to organize: a physician-led alliance that pools clinical voices so doctors help write the rules for AI in medicine instead of receiving them. Its value is collective weight. An individual asking for a governance seat is easy to defer, but a physician backed by an organized alliance carries the standing to demand one, and to hold it.

The Part of Medicine No Model Can Take

The work that does not commoditize is the work that protects you. A model can read the scan, draft the note, and surface the differential. It cannot sit with a frightened patient and decide which truth they are ready to hear today. It cannot earn the trust that makes a noncompliant diabetic actually change. It cannot weigh a frail eighty-year-old’s stated wish to go home against the textbook plan and know that honoring the wish is the better medicine.

The patient relationship is not a soft skill that survives automation by luck. It is the appreciating asset. As the routine cognition gets cheap, a clinical physician’s presence, judgment under real uncertainty, and the moral weight of a recommendation get scarce, and scarce things rise in value. The physician who treats the AI as a way to reclaim minutes, then spends those minutes in the room rather than on the next click, is investing in exactly the capacity the technology cannot reproduce. Patients can already tell the difference between a doctor who is present and one who is performing documentation while half-listening, the kind of paperwork load the latest MGMA data shows is still winning. Soon that difference will be the whole of what they are paying a human for. I made the longer version of this argument in my earlier piece on antifragility in the Neuroeconomy, and the C-suite story is the same argument seen from the org chart.

So protect it deliberately. The reclaimed hour from an AI scribe is not a productivity dividend for the system to harvest. It is the hour you give back to the bedside, and it is the strongest argument you have that medicine still needs you in it.

How to Be Valuable to the AI Chief

You do not have to become a data scientist. You do have to become conversant, because the executives reshaping your practice respect fluency, and they make room for the clinicians who have it.

Learn the vocabulary that governs these decisions. Go beyond the difference between sensitivity and specificity to understand the receiver-operating characteristic curve and its relation to precision, because you need to be able to challenge a vendor’s headline number: a 95% sensitive tool with a poor false-positive rate will flood you with unnecessary alerts. Understand what model drift means, that a model trained on last year’s patients degrades on this year’s, which is why revalidation schedules matter. Know that the Coalition for Health AI, known as CHAI, has published governance frameworks that your system is likely already citing, and that being able to reference them puts you on equal footing in the meeting.

Understand the economics, too. Bain and KLAS have documented AI moving from pilot projects to production across health systems, and finance leaders increasingly name the chief AI role as the most important emerging seat in the C-suite. When you can connect a clinical concern to the financial and operational case the executive is accountable for, you stop being the doctor who resists and become the partner who makes the deployment safe enough to defend to the board. That is the physician an AI Chief actually wants in the room.

The AHA framed the moment precisely: health systems are no longer asking whether technology gets a seat at the table, but what kind of seat and how much authority it should hold. That sentence is about the executives. It should be about us, too. The authority is being apportioned right now. Florida physicians who learn the language, claim the governance seats, and guard the part of medicine no model can touch will not be reshaped by this C-suite. They will help shape it.

Frequently Asked Questions

What is a Chief AI Officer in a hospital, and what do they do?

A Chief AI Officer, sometimes titled Chief Health AI Officer or Chief Data and Artificial Intelligence Officer, is a C-suite executive who owns a health system’s AI strategy, governance, and safety, usually reporting directly to the CEO. They decide which AI tools the system buys, how they are configured, and what outcomes those tools are tuned to improve, which makes the role a de facto author of clinical policy.

Are Florida hospitals hiring Chief AI Officers?

The trend is national and accelerating, reaching systems from Cleveland Clinic to rural networks like Sanford Health, and Florida’s large systems are moving in the same direction. Given Florida’s aging population and uneven physician distribution, the configuration choices these executives make will land especially hard on the state’s patients, which is why Florida physicians have a direct stake in who holds the role.

How can physicians keep influence as AI executives gain power?

Claim a real seat on the AI governance committee, ask in writing what each new tool is optimized for and who measures it, and audit a model you already use for its validation data and false-positive rate. Back individual requests with the Florida Medical Association so a single complaint becomes a negotiation the system cannot route around.

Will AI replace the physician-patient relationship?

No. Models can read images, draft notes, and surface differentials, but they cannot hold a frightened patient’s trust, weigh a patient’s stated wishes against the textbook, or carry the moral weight of a recommendation. As routine cognition becomes cheap, that human work becomes scarcer and more valuable, which is why physicians should reinvest AI-reclaimed time into the bedside rather than let the system harvest it for throughput.

What should a physician learn to work well with a Chief AI Officer?

Become conversant rather than expert. Understand sensitivity versus specificity, the ROC curve and precision, model drift and revalidation, and the governance frameworks from the Coalition for Health AI that your system likely already references. Connect clinical concerns to the financial and operational case the executive is accountable for, and you become the partner who makes a deployment defensible rather than the obstacle to it.

Sources

• American Hospital Association, “The C-Suite Rewrites the Org Chart: How Health Systems Are Elevating Technology Leadership” (February 2026).
• Cleveland Clinic Newsroom, “Cleveland Clinic Names First Chief AI Officer” (July 2024).
• American Medical Association, “2 in 3 physicians are using health AI—up 78% from 2023” (Augmented Intelligence Research, 2025).
• American Medical Association, “Do health care organizations need a chief AI officer?”
• Healthcare Dive, “CHAI releases AI governance guidance for health systems.”
• Bain & Company, “Healthcare IT Investment: AI Moves from Pilot to Production.”

Hero image: illustration created with AI image tools.