Her Two AI Holdouts Were the Youngest Providers in the Building

By Jill Steeley, founder of Steeley Strategic Solutions, former FQHC CEO, and host of The Community Health Collective podcast. I help health center leaders build organizations that are financially profitable, sustainable and true to their mission.

Highlights:

• A CEO rolled ambient scribes out to 50 providers and got exactly two holdouts - both her youngest providers

• The objections weren't fear or confusion. They were ethics and environmental cost.

• There are three kinds of "no," and standard change management only addresses two of them

• A capability no shrinks with support; a values no hardens when it's ignored

• Four moves that work on a values objection, none of which are another training session

• Your loudest ethical objector is pressure-testing your rollout for free

• The real lesson is bigger than AI: stop predicting who will resist

Picture the person in your building who's going to fight you hardest on implementing AI.

You already have a face in your head. I'd bet money on the general shape of it - long-tenured, later career, still likes paper charts. Every CEO I say this to has an immediate, confident answer.

Hold that picture, because a CEO in one of our CEO Connect Bootcamp sessions had the exact same one. And she was wrong.

Why did the youngest providers refuse the AI scribes?

She rolled ambient scribes out across her health center. Fifty providers. The AI listens to the visit, drafts the note, the provider reviews it, signs, and locks the chart. It's a big lift - it took months of planning, real money to rollout, and a full implementation project.

She braced for the fight. She had a list in her head of exactly who was going to give her trouble: the providers who've been doing this thirty-plus years, the ones who would go back to paper charts if anyone let them. Her communication plan, her training plan, her extra support - all of it aimed at them.

She got two holdouts out of fifty.

Neither one was over 40. Both were her youngest providers - digital natives who have never in their professional lives known a world without a smartphone, a tablet, and an electronic health record.

And here's what caught the attention of everyone in that Bootcamp room: they weren't afraid of it. They weren't confused by it. They didn't need another training. They understood the technology better than almost anyone else in the building.

They objected on ethics. And they objected on environmental cost.

Meanwhile, her 62-year-old doctors adopted in week one. Several of them love it. Of course they do - they're the ones who've been staying until eight at night finishing notes for twenty years. You hand that person two hours of their evening back and they will not ask you a single philosophical question.

What are the three kinds of "no" in a technology rollout?

Here's what I want you to take from this episode: you cannot answer a values objection with a training session.

When somebody won't adopt, it's almost always one of three things. From the outside, they look identical.

1. "I can't." — the capability no. They don't know how. The workflow doesn't make sense. The integration is clunky. Nobody showed them the right way to use it. This one is real, it is by far the most common, and it genuinely is a training problem.

2. "I don't trust it yet." — the evidence no. They're worried the note will be wrong and their name is on it. They're worried about the liability and their license. They want to see it work before they hand over something they're personally accountable for. Start with a trusted colleague and let them see the benefits and results first hand.

3. "I don't think we should." — the values no. This is a different animal entirely. It does not respond to either of the first two fixes. You can train someone until you're blue in the face and hand them 100% accuracy data, and the answer will still be no - because you haven't touched the actual objection.

I teach change management, and here's the trap: nearly every change-management playbook you've been handed is built for the first two. So you keep offering more support, more office hours, more lunch-and-learns. The person keeps saying no. And you start to think they're being difficult.

They're not being difficult. You're answering a question they never asked.

Why does ignoring a values objection make it worse?

This is the practical part, and it's the reason the distinction matters.

A capability no and an evidence no both get smaller over time if you support them well.

A values no gets bigger if you ignore it.

It hardens. The person's thinking calcifies around it. They pick up allies - or they teach their allies to believe the same thing. And critically, the person holding a values no usually gets quieter, not louder. Which means by the time you notice it, it's already spread through your organization.

So before you spend one more dollar on training, ask which no you're actually looking at. The only way to find out is to ask the person directly.

What were the ethical and environmental objections, exactly?

I want to lay these out, because I don't think most of us have sat with them.

The ethics questions: What happens to the audio from that exam room? Is it stored, and for how long? Is a patient's conversation about their addiction, their abuse history, or their immigration status being used to train somebody's model? Does the patient actually understand what they're consenting to when the front desk hands them a form? And is a "yes" from a patient who has nowhere else to go a real yes?

Those are not naive questions. Those are the exact questions your compliance officer should be asking. Those are the questions your board is going to ask you in six months.

The environmental argument: that the data centers running these models consume enormous amounts of electricity and water, that the buildout is happening fast, and that health care of all industries should think hard about the downstream health effects of what it's fueling.

I'll be honest with you - I hear this argument in my own house. My son makes it to me. He's not anti-technology. He just thinks the people using it should have to look at the whole cost, not only the part that shows up on the invoice.

Do I agree with every piece of it? No. Do I think it's a reason to stop? No. But I've stopped treating it as a fringe position, because it isn't one. It's a generational position - and that generation is exactly who you're trying to recruit and retain right now.

Here's the thing. The younger clinicians in your building are not resistant to technology. They're resistant to unexamined technology. That's a completely different problem, and frankly, it's a healthier one to have.

What are the four things to do with a values objection?

You've diagnosed it. It's a values no. Here's what to do, and none of it is another training session.

1. Name it out loud, in the room. Say the words: "I don't think this is a training issue. I think you have an objection to the technology itself. Am I right?" That question alone changes the temperature. Most people holding a values no have never been asked directly - they've just been managed around. Being asked is disarming, it's respectful, and it gets you the truth.

2. Go get the actual answers. Not vendor marketing - answers. Take their questions to the vendor in writing: Is our patient audio retained, and for how long? Is our data used to train your models, and can we contractually prohibit it? What exactly does the patient consent language say, and can we improve it? What do you publish about energy use? Then bring the answers back to the objector, including the ones that don't support your position. Half the time the answers are better than they feared and the objection shrinks on its own. The other half, you just found a real problem in your contract before it became a big one.

3. Give a real opt-out with a real boundary. Two out of fifty is not a crisis. Let them document the old way. Be clear about what that costs and where the boundary is: this is available now, it may not be forever, and here is the date we will revisit it. A genuine revisit date isn't weakness - it's what keeps the objector from becoming a toxic influence, and it's what keeps the other 48 moving.

4. Don't make them the villain. The fastest way to turn two holdouts into twelve is to talk about them like they're the obstacle. The minute your leadership team starts rolling their eyes about "the two who won't get on board," you've told everyone else that raising a concern here is a career risk. And then you stop hearing concerns. You do not stop having them.

Why your objector is an asset, not a personnel problem

Here's the reframe I most want you to take.

Those two providers were doing free work for that health center, and nobody had noticed.

Every question they raised about consent and data retention is a question a patient will eventually ask at your front desk. A reporter could ask it. A site visitor could ask it. A board member absolutely will.

You have people on staff pressure-testing your rollout out of conscience, on their own time. That is not a personnel problem. That's an early-warning system, and it costs you nothing.

So use them. Put your loudest ethical objector on the committee implementing the technology and writing your consent language. Ask the one worried about the data to help you review the vendor contract. You'll get a better policy - and this is the part most people don't expect, you'll often get the objector too. Not always. But often. Because most people who object on principle aren't trying to stop you. They're trying to be heard before you hit go.

And the ones who don't convert? At least you'll know exactly where they stand, instead of getting a nod in a meeting from someone who's quietly planning to work around you.

The lesson that's bigger than AI

Her rollout went great. Forty-eight of fifty. By any measure, a win.

But she spent months preparing for a fight that never happened, and she had no plan at all for the one she got.

The assumption cost her more than the resistance did.

She planned around a stereotype instead of around her actual people. And the two things she was completely unprepared for turned out to be the only two things she had to deal with.

So don't take "young people are the new holdouts" from this. That's just a different stereotype, and it'll be wrong for someone too.

The lesson is: stop predicting who will resist. Go ask.

Because change management built on who you assume will struggle isn't change management. It's a guess with a project plan wrapped around it.

Your one thing this week

Pick whatever you're rolling out right now. AI, a new EHR module, a new workflow, a new policy - it doesn't matter.

Find the people who haven't come along yet.

Ask each one a single question: is this a can't, a don't trust it yet, or a shouldn't?

Don't defend anything. Don't sell. Just listen and sort them into three buckets. That fifteen-minute conversation will tell you more about your rollout than any survey you could send.

And if you find a values no, you now know it needs a completely different conversation — not more training.

This is exactly the kind of thing we work through in the FQHC CEO Connect Bootcamp. This story didn't come from a case study I found somewhere. It came from a CEO sitting in one of our sessions with a room full of peers, saying the honest version out loud and getting real answers back from people who had just been through it. We had another CEO in that same room who has successfully implemented ambient scribes and was willing to share his data - that was an eye-opener for the whole group. That's the part you can't get from my podcast or a webinar.

🎧 Listen to Episode 40 of The Community Health Collective

📅 Or reach out and we'll get on a call to see whether the next cohort is the right fit: www.fqhc-ceo.com

This work is hard, and it's getting harder. But you're capable of more than you think - and you're not doing this alone.

About the Author

Jill Steeley is the host of the Community Health Collective Podcast and an executive coach to leaders across community health centers, FQHCs, and mission-driven healthcare organizations. After two decades inside the healthcare leadership world and close to 250 healthcare leaders coached and mentored, she helps healthcare executives build the leadership skills they were never formally taught - and helps full leadership teams shift culture together rather than one leader at a time.

Learn more at jillsteeley.com.