AI Is Finally Paying Off in Medicine. The Real Question Is Who Gets the Check.

By Shane Cole, MD

Fortune ran a piece this week about something most of us in the trenches already feel. AI is starting to deliver real returns in health care. Clinicians using it well are seeing roughly eight more patients a week. Administrative load is dropping. Patient satisfaction is climbing. More than half of patients who interact with AI report that they actually like it.

That last part might be the most surprising line in the whole report. A year ago, most patients were nervous about AI in their care. Now they are getting comfortable with it, mostly because the version of AI they are running into is the one that makes their experience smoother. Faster check-in. Cleaner notes. Less time waiting for someone to type instead of looking at them.

The Philips Future Health Index is calling this the AI dividend. They are right. The dividend is real.

The fight, which the article gets to about halfway through, is over who keeps it.

I have spent more than two decades in emergency medicine, and the last ten years building and running my own freestanding ERs. So I have watched this movie from both seats: the doc on shift, and the owner signing the contracts. Here is the part every emergency physician should pay attention to. The last time medicine got an efficiency boost, telehealth, the insurers ate most of it. We became more efficient, and the response from payers was not to reward us. The response was to lower reimbursement and shrink margins, because in their math, you did less work per encounter. The clinicians who delivered the efficiency did not get to keep the dividend. The system did.

There is no reason to assume AI will play out any differently unless physicians actively make sure it does.

The leaders quoted in the Fortune piece are saying the right things. AI is changing how we care for people. It needs to be an enterprise strategy, not an IT project. It should expand the moments of empathy between clinician and patient, not replace them. All true. All worth saying.

But none of that decides who profits.

That decision is happening in three places.

First, at the contracting table. Insurers will quietly use AI gains as a reason to argue that the work is now easier and reimbursement should reflect that. If physician groups do not show up with their own data, their own efficiency story, and their own position on the value AI is creating, the rate cuts will arrive without resistance. I have sat across from payers. They are never going to make that argument for you.

Second, in the technology stack itself. Hospitals and large platforms are racing to own the AI layer. If physicians do not have a hand in choosing the tools, training the models, and shaping the workflows, we will spend the next decade carrying systems that were built to make administrators more efficient, not us. In our ERs, we get to make those choices ourselves, and we intend to use that freedom. Our plan is for every facility we operate to be the most technologically advanced emergency room in its service area. Not as a marketing line, but because when physicians pick the tools, the tools serve the bedside.

Third, in ownership of the practice itself. This is the one I know best. The most important shift AI is creating is not faster charting. It is the ability for a small, well organized clinical team to do the work of a much larger one. That changes what is possible for physician owned and physician led groups. The leverage that used to require a hospital system can increasingly be built by a focused independent group with the right tools, the right partners, and the right operating discipline.

I drew up the plan for our first freestanding ER in 2013 and opened the doors at the end of 2015. We have three locations running and five more coming. I am not telling you that to brag. I am telling you because ownership is the difference between earning the dividend and being handed whatever is left of it. When I was an employee, efficiency gains belonged to somebody else. As an owner, when our team gets better, faster, and smarter, the people who built that improvement keep the value, including the physicians working the night shifts and absorbing the risk.

This is why CAM exists. We believe the next era of emergency medicine has to be physician aligned. Not because that sounds nice in a brochure, but because if the AI dividend lands in the wrong pocket, the people delivering care, the ones absorbing the risk, the night shifts, the patient outcomes, end up carrying more weight for less money. Again.

The good news is the moment is wide open. Patients trust it more than they did. Clinicians are seeing real benefits in their day. Leadership at major systems is, at least publicly, saying the right things. There is real momentum, and there is still room for physicians to claim our share rather than have it allocated for us.

The hard truth is that no one is going to hand us this dividend. Not the insurers. Not the platforms. Not the systems. We will have to build the structures that let us keep it. I built mine because I wanted a career where my family came first, my health came second, and the financial security followed from owning what I built rather than renting myself out by the shift. AI just made that path wider for the next physician willing to walk it.

That is the work. Not just learning the tools. Building the businesses that own them.