What Dentists Want AI to Do First

In June, the ADA's Health Policy Institute asked 552 dentists in private practice a straightforward question: for which of the following tasks do you use AI in your practice?

Just under half said they already use it for at least one thing. 43.3 percent currently use AI somewhere. Another 26.4 percent do not but plan to. And 30.3 percent do not and do not plan to. Those three numbers are shares of everyone who answered, and they add to a hundred.

That much has been covered. What has not been covered is the second question HPI asked, and the second question is the interesting one.

First, where dentists actually are

Broken out by task, current use is concentrated where you would expect. Imaging and diagnostics leads at 22.8 percent. Insurance verification is at 13.6 percent, explaining clinical findings to patients at 13.2 percent, social media at 10.7 percent. Receptionist and front-desk check-in tools sit at 10.1 percent, appointment scheduling at 9.2 percent. Treatment recommendations are last, with fewer than 5 percent of dentists using AI for them at all.

Now look at intent. Among all 552 respondents, the tasks the most dentists say they do not use AI for but plan to are charting and note taking at 34.8 percent, insurance verification at 32.6 percent, billing and claims at 29.7 percent, and practice analytics at 27.2 percent.

Front-desk check-in and appointment scheduling are both at 19.6 percent. Near the bottom of that list. Fewer dentists plan to bring AI to the front desk than plan to bring it to charting, insurance, billing, or analytics.

If you stopped reading there you would conclude the front office is an afterthought.

Then HPI asked the people who plan to adopt when they plan to start

The second question went only to the dentists who said they do not use AI for a given task but plan to. It asked when. The only two answers were "in 2026" and "2027 or later."

Front-desk check-in tools came first. Of the dentists who intend to use AI there, 32.4 percent say they will start this year rather than next year or after.

Appointment scheduling came second, at 29.6 percent.

Practice analytics is third at 29.3, charting fourth at 29.0. Imaging and diagnostics, the task the most dentists already use AI for, is near the bottom of the timing list at 19.3 percent. Treatment recommendations are at 16.9. Explaining clinical findings to patients is last at 14.7.

So the two lists run close to opposite. The front office is the thing the fewest dentists plan to do, and it is the thing that the dentists who do plan it intend to start soonest.

That is a real finding and as far as I can tell nobody has reported it. Every piece of coverage on this survey has stopped at the first question.

Read that carefully before you build anything on it

A few things this does not say, and they matter.

It does not say front-desk AI is more popular. It is less popular by volume. Charting has nearly twice the intent behind it.

It does not say anything worked. HPI measured use, intent, timing, and how dentists feel. It published nothing about results, hours saved, or whether anyone who started was glad they did. There is no ADA figure on the return from any of this, and if you see one quoted, it was invented.

And the sample is what it is. HPI says so plainly: compared to the general population of practicing dentists, this quarter's respondents skewed toward ADA members, general practitioners, practice owners, working solo or in small groups, mid to late career, white, and urban. That skew happens to run toward owner-operators of small practices, which is most of the people reading this, but it is a skew and it should be named.

Why the soonest thing is the front office

HPI offered its own reading, and it is worth quoting because it comes from the people who ran the survey. Writing about these results, HPI said the biggest opportunities in the near term seem focused on easing administrative burdens, and connected that directly to staffing: dental practices are struggling with ongoing staffing shortages, dentists are spending more of their work week on non-clinical tasks, and some dentists may view AI tools as a workaround to ongoing staffing problems.

That explains the timing better than anything else. The front office is not where the most dentists are headed. It is where the ones who are headed there feel the pressure now.

There is a second explanation sitting in the same survey, in the dentists' own words. HPI asked how they feel about the expanded use of AI in dentistry and coded the open-ended answers. The single most common reaction was general optimism about the potential, at 21.5 percent. The second, at 17.7 percent, was that AI is a tool and should not replace clinical judgement. One respondent put it this way: "AI is a fantastic tool if used responsibly, it should be an aid to a job, not the only source of competency. The operator should always be the one in command of the outcome."

That is not resistance to AI. It is a bar, and it is a higher bar for clinical work than for anything else. Charting has more intent behind it and a longer runway because the standard a dentist holds it to is a clinical standard.

The front office has a lower bar, and the reason is not that it matters less. It is that the work is already defined and already measured. A call was answered or it was not. An appointment was booked or the patient hung up. A message went out or it sat. There is no diagnostic judgement to defer to, and there is nothing to argue about after the fact. A practice can tell within a week whether the phones are being handled better than they were.

That is why it is the soonest thing.

What that means if you are the one deciding

If you are weighing where to start, the survey is not telling you to do the front desk. It is telling you that the dentists who picked the front desk are moving now, and it is worth understanding why they can.

The honest test is whether you can tell if it worked. Pick the task where a bad outcome is visible to you within a couple of weeks and where the standard is something your team already agrees on. For most practices that is the phone, because everyone in the building already knows what a missed call is.

The harder tasks are not off the table. They are just harder to evaluate, which is exactly what the timing answers show dentists already understand.

Laine is the piece of what we build that sits on the phone. It answers the calls the practice cannot get to, including after hours and on weekends, and it books into the schedule with write-back to Dentrix, Dentrix Ascend, Open Dental, or CareStack. If you want to hear what a patient hears before you connect it to anything, you can. There is no long-term contract and no onboarding fee.

Deren Flesher, DDS. Founder, AiroDental.

---

Figures throughout are from the ADA Health Policy Institute's State of the U.S. Dental Economy, Second Quarter 2026 Update, based on 552 dentists in private practice surveyed from June 15, 2026. The three-way use figures and the per-task figures are shares of all respondents. The timing figures are shares of the dentists who said they plan to use AI for that specific task.