Is AI really being used to deny patients healthcare?

Yes, in a limited but real way. In mid January 2026, Medicare launched a prior authorization pilot called the WISeR Model in six states, Oklahoma, Arizona, New Jersey, Ohio, Texas and Washington, that uses AI from third party tech firms to review certain procedures like epidural injections before they are approved. KFF Health News reported delays and denials, including nearly 100 patients in one Washington health system waiting on epidural injections due to WISeR related holdups. An American Medical Association survey found 61 percent of physicians worry AI will increase denials of necessary care. So the fear patients hear about is grounded in real reporting, even though it involves insurers and Medicare, not the friendly AI a local practice might use to answer the phone.

Will using an AI receptionist make patients trust my practice less?

Not if you use it for the right job. Patients react badly to AI that gatekeeps or decides against them, like an insurer bot that denies a claim. They react well to AI that serves them, like a system that answers the phone at 9pm, texts back a missed call in seconds, and books their appointment fast. The trust comes from three things: the AI actually helps the patient rather than blocking them, you are honest that it is an assistant, and a real human is always one step away. Used that way, an AI receptionist tends to raise trust because the patient finally gets a fast answer instead of voicemail.

Should I tell patients when they are talking to AI?

Yes. Transparency is cheap and it protects trust. You do not need a legal disclaimer, just a plain line like this is our virtual assistant, I can book you or connect you to the team. Public openness to AI in healthcare fell from 52 percent in 2024 to 42 percent in 2026, and a Pew Research Center survey found 60 percent of Americans would feel uncomfortable if their provider relied on AI. In that climate, hiding the fact that a patient is talking to a bot is the fastest way to lose them when they figure it out. Being upfront, and making a human easy to reach, is what keeps the trust.

What kinds of AI do patients actually accept from a practice?

Patients accept AI that removes friction for them and rejects AI that stands between them and care. Accepted: instant answers to common questions, after hours booking, missed call text backs, appointment reminders, help finding your hours or address. Resisted: anything that feels like it is judging them, denying them, or replacing the doctor's clinical judgment. The simple test is whether the AI works for the patient or against them. Keep your AI firmly on the patient's side, doing the boring, fast, helpful work, and leave every clinical and human moment to your team.

How do I keep the human touch while using AI in my practice?

Draw a clear line. Let AI handle the repetitive front desk load that patients hate waiting on, answering the phone, capturing missed calls, booking, reminding, and let humans own every moment that carries care or emotion, the clinical visit, a worried patient, a complaint, a bad diagnosis. Always give patients a fast, obvious way to reach a person, and never let the AI pretend to be a doctor or make a medical decision. When the routine is automated, your team actually has more time for the human part, which is the opposite of what patients fear when they hear healthcare AI.

Healthcare AI and Patient Trust in 2026

A clinician checks a patient's blood pressure by hand, the human care that patients trust more than faceless healthcare AI — The thing patients are afraid of losing: a real person, paying attention. That is your advantage. Photo via Pexels.

A headline went around this month that made a lot of patients nervous. BBC Science Focus ran it plainly: AI is now deciding who gets healthcare in six US states, and it is already harming patients. Whatever you think of the framing, millions of people read some version of that this week. And once a patient reads that AI can be used to deny their care, they carry that feeling into every place healthcare AI shows up next. Including yours.

So let us talk about what really happened, because the details matter, and then about the part that actually affects your schedule: the trust your practice is either building or losing every time a patient meets a bit of AI at your front door.

What is actually happening in those six states

In mid January 2026, Medicare launched a pilot called the WISeR Model in six states, Oklahoma, Arizona, New Jersey, Ohio, Texas and Washington. It uses AI from outside tech companies to review certain procedures before they are approved, things like epidural injections, kyphoplasty and skin substitutes. The goal on paper was to cut waste. The result on the ground, according to reporting from KFF Health News, has been delays, glitches and unexpected denials.

The specifics are hard to read. One Washington health system had nearly 100 patients waiting on epidural injections earlier this year because of WISeR related holdups. A 65 year old Oklahoma cattle rancher who had driven two and a half hours for a routine injection for years suddenly needed a whole extra trip, one appointment just to fill out a piece of paper. Doctors quoted in the reporting called the rollout horrendous. Members of Congress from the affected states are trying to repeal it.

This is prior authorization by insurers and Medicare, not a doctor's office using a chatbot. But patients do not sort it that neatly. They hear two words together, healthcare and AI, and they hear that it left a rancher in pain. The American Medical Association found that 61 percent of physicians already worry AI will increase denials of care patients need. When even doctors are saying that out loud, patients believe it.

Why this is your problem, even though it is not your AI

Here is the uncomfortable part. The insurer's AI is the villain in that story, and your practice never appears in it. But you inherit the mood anyway. Trust does not travel by category. A patient who just read that AI harmed someone does not think, well, that was a Medicare contractor, surely the bot at my dermatologist is different. They think AI in healthcare is something done to me, not for me. And they bring that suspicion to your booking page.

The numbers back this up, and they are moving the wrong way. A Pew Research Center survey found 60 percent of Americans would feel uncomfortable if their own provider relied on AI. More recent tracking shows public openness to AI in healthcare fell from 52 percent in 2024 to 42 percent in 2026, even as more practices adopt it. And separate research on patients calling their providers found that 90 percent say they want to reach a real person, not a machine. The gap between how fast practices are adopting AI and how uneasy patients feel about it is widening, not closing.

So if you are adding AI to your front desk, you are doing it into a headwind. That is not a reason to avoid it. It is a reason to do it in a way that reads to the patient as help, not gatekeeping. Because the very same patient who fears the insurer's denial bot will happily let a bot book their appointment at 9pm, if it clearly works for them.

Patients now sort AI into two buckets

This is the whole insight, so sit with it. In the patient's mind, healthcare AI now splits into two very different things:

AI that decides against them. The insurer bot that denies the claim, the algorithm that delays the injection, the system that stands between them and their doctor. This is the AI in the headlines, and patients hate it.
AI that serves them. The assistant that answers the phone instantly, texts back a missed call in seconds, and finds them an appointment this week instead of leaving them on hold. Patients barely think of this as AI at all. They think of it as finally getting a fast answer.

The trap is that a nervous patient can mistake the second kind for the first if you set it up carelessly. A cold, evasive bot that hides what it is, dead ends every question, and never lets you reach a person feels like the denial machine, even if all it does is book appointments. Get the design wrong and helpful AI still triggers the fear. Get it right and it becomes the reason a patient trusts you more, not less.

How to use AI so it builds trust instead of fear

The good news is that the rules for staying on the right side of this are simple, and none of them are technical. They are about respect.

Be honest that it is an assistant. You do not need a legal disclaimer. One plain line does it: this is our virtual assistant, I can book you or connect you to the team. Hiding the bot is the fastest way to lose a patient the moment they realize. We went deeper on this in whether you should tell patients when AI answers, and the short version is always yes.

Keep a human one tap away. The single biggest fear in that WISeR story was being trapped with no person to appeal to. Your AI should do the opposite. At any moment, a patient should be able to reach a real member of your team, easily and fast. AI that helps you get to a human is trusted. AI that replaces the human and blocks the exit is feared.

Point it only at jobs that help the patient. Answering common questions, booking, missed call text backs, reminders, hours and directions. Never let it make a clinical call, judge a patient, or stand between someone and care. The test is dead simple: does this AI work for the patient or against them? Keep it firmly on their side. We laid out the full framework in our guide to using AI without losing patient trust.

Make it fast and warm, not robotic. Speed is the thing patients actually notice. A missed call answered by text in ten seconds beats a voicemail they wait two days on. When the AI is quick, helpful and clearly hands off to a person, patients experience it as your practice caring enough to answer, which is exactly the human feeling the headlines say AI is killing.

Our honest take

We will say the blunt thing out loud. A lot of the fear about healthcare AI is earned. When AI is pointed at patients to save someone else money, to deny, delay and cut, it deserves the backlash it is getting, and we would never build that. But the reaction to that story is pushing some good practices to avoid AI entirely, and that is the wrong lesson. The rancher in Oklahoma was not harmed by AI answering a phone. He was harmed by AI used as a wall.

The practices that will win the next couple of years are the ones that draw that line clearly and out loud. Use AI to remove the friction patients hate, the hold music, the after hours voicemail, the three day callback, and put your humans on everything that carries care. Done that way, AI does not make you feel less human to patients. It frees your team to be more human, because they are no longer buried in the phones. That contrast, a practice that uses technology for the patient while the insurance world uses it against them, is one of the strongest trust signals you have right now. Use it.

How EtherealMinds sets this up

When we build a practice's patient acquisition system, the AI is aimed squarely at the patient's side of the table. Our AI receptionist answers every call and message, day or night, tells people plainly that it is a virtual assistant, books them in a few taps, and hands off to your team the instant a real person is needed. It never pretends to be a doctor and it never plays gatekeeper. For a field like pain management, where the WISeR delays are hitting hardest and patients are already on edge about being stuck in a queue, that fast, honest, human backed front door is worth a lot.

We pair it with a website built to convert that makes booking obvious and quick, so the whole path from I need help to I have an appointment feels like a practice that answers, not a system that stalls. If you have been holding off on AI because you do not want to feel like the insurance company, good instinct. The answer is not to skip it. It is to build it so patients can tell, immediately, whose side it is on.

Use AI to win patient trust, not lose it

Book a free strategy call. We will look at how patients reach your practice today, where they hit voicemail or a wall, and show you how to add AI that answers fast and stays honest, with a real person always one tap away. Clear plan, real numbers, no jargon, no pressure.

Book a free strategy call →

AI Is Denying Care in 6 States. Patients Now Eye Your AI Too