What is the Front Door to Healthcare?

For decades, the path to healthcare started the same way: call the doctor's office, wait for an appointment, show up, and hope for answers.

That model assumed patients had no alternative. For most of modern medicine, they didn't. The doctor's office was the front door because it was the only door.

That's no longer true.

In the first three months of 2026, five of the largest technology companies launched dedicated health AI products. Amazon embedded Health AI where 200 million people already order household goods. Google processes over a billion health queries every day. OpenAI's ChatGPT Health handles 230 million health-related conversations per week. Anthropic and CVS followed with their own offerings.

And consumers are responding. One in three U.S. adults now uses AI chatbots for health information, double the rate from just a year ago. Most aren't using tools built by health systems. They're using ChatGPT.

The shift is behavioral, not just technological

The important thing isn't that these tools exist. It's that patients prefer them.

Seventy percent of health-related AI queries happen after hours, when clinics are closed. Somewhere between 42% and 58% of people who consult AI for a health question never follow up with a provider at all. They get an answer, or at least something that feels like an answer, and move on.

For a rural health system like Ballad, where the nearest specialist may be two hours away, the appeal of instant AI is obvious. Our patients aren't choosing AI over us because they don't trust us. They're choosing it because it meets them where they are at a cost approaching free.

What tech companies understand

These companies aren't building chatbots for the sake of chatbots. They're positioning AI as the wedge to own the upstream relationship, the moment a person first wonders if something is wrong.

The healthcare chatbot market is projected to grow from $2.4 billion to $12.6 billion by 2034. But the real prize isn't chatbot revenue. It's becoming the default place people turn when they have a health question. Whoever owns that moment shapes what happens next: which provider they see, which pharmacy fills their prescription, which health plan they trust.

For most of the 20th century, that moment belonged to the doctor. For most of the 21st, it belonged to Google Search. Now it's moving again, toward conversational AI that feels more personal, more responsive, and more available than anything that came before.

What do we actually have?

It's tempting to say health systems have something these companies can't replicate. But let's look at what's changed.

Data access is no longer a moat. Thanks to interoperability rules and health information exchanges, tech companies can now compile your medical records just like we can. Amazon's Health AI already pulls diagnoses, medications, and history before responding to your questions. You can load medical history into ChatGPT and Claude as well. The idea that we hold the full picture, and they don't, is increasingly outdated.

Time isn't on our side either. AI can converse with you for as long as you want. A typical doctor visit is 15 minutes, and some of that goes to administrative tasks. The chatbot is infinitely patient. It doesn't have a waiting room full of other patients.

Wellness is a blind spot. These platforms move seamlessly between wellness and illness, from sleep optimization to symptom checking to medication questions. Most physicians aren't trained in wellness. They are trained to diagnose and treat disease. The consumer doesn't draw a line between "feeling better" and "not being sick," but our entire system is built around that distinction.

Knowledge is no longer proprietary. The clinical decision support systems that help doctors make better decisions are built on the same AI models available to consumers. The gap between what a physician knows and what a well-prompted chatbot can surface is shrinking fast.

So what do we actually have?

We have the power of the pen, the legal authority to prescribe treatments, order tests, and make medical decisions. But that's being unbundled too. Companies like Ro, Hims, and Amazon Clinic have built entire businesses around lightweight prescribing for common conditions. For a growing list of problems, you don't need a health system to get a prescription.

We have bricks and mortar, physical facilities and trained experts for the things you can't do remotely. Surgery. Imaging. Emergency care. Procedures. Labor and delivery. Complex inpatient stays. When something serious happens, you still need a hospital.

But that's a narrower value proposition than we've historically claimed. It means we're essential for high-acuity care, and increasingly optional for everything else.

The question for health systems

In a knowledge-based industry where knowledge is being commoditized, what's left?

If AI can answer the question, compile the record, and triage the need, and if someone else can write the prescription, what exactly is the role of the health system before you actually need a procedure?

We don't have a complete playbook for this. But we think the path forward starts with identifying the trend, and being willing to explore answers we haven't considered yet.