One in Three Americans Now Uses AI for Health Information. Your Budget Still Doesn't Know It.

A doubling happened while most healthcare marketing teams were looking the other way.

One year ago, 16% of Americans used AI chatbots for health information. That number is now 32%. Not a gradual adoption curve. A behavior shift that already occurred while healthcare marketing budgets were still calibrated for Google.

The Rock Health Consumer Adoption Survey, fielded with 8,000 U.S. adults, puts hard numbers on something we have been tracking in real client data for over a year: patients are making care decisions in a channel most health systems do not measure, cannot influence, and are not prepared for.

32%  of Americans now use AI chatbots for health information — up from 16% one year ago

73%  of AI health users went to ChatGPT first. Provider-built chatbots: 5%.

81%  of AI health users took a follow-on action after their query

40%  consulted a provider after an AI interaction

64%  use AI for health questions weekly or more

Where Patients Actually Went

Of the one in three Americans who used AI for health information, 73% went to ChatGPT first. Provider-built chatbots captured 5% of that traffic. Payer-built tools got 4%.

Health systems have invested heavily in digital front doors over the past decade. Patient portals. Appointment scheduling tools. Symptom checkers. The Rock Health data tells us patients walked around all of it. Not because the tools were bad. Because ChatGPT was faster, more available, and did not require them to already know which health system to go to.

That last point deserves to sit for a moment. The patient is forming a preference before they reach your digital front door. By the time they show up on your website, the decision may already be made.

AI Is Not Informing Patients. It Is Routing Them.

The 81% follow-on action rate is the number most healthcare marketers are underreacting to. When 4 out of 5 AI health users take action after their query, and 40% go on to consult a provider, AI is not functioning as a passive information resource. It is functioning as a referral layer. It influences which specialist gets the call, which health system gets the appointment request, and which program gets considered. The organizations cited favorably in that AI response are capturing volume. The ones not cited are invisible. And unlike a missed paid search impression, there is no log of the miss. The patient simply does not show up.

Your CRM has no record of these patients because they never entered your funnel. Your analytics dashboard lacks a column for sessions that never happened. The only way you know there is a problem is when volume numbers start moving in a direction your other data cannot explain.

Your Highest-Value Patients Are the Most Affected

The demographic profile of AI health users in the Rock Health data should get the attention of any CMO focused on margin and lifetime patient value. AI users are healthcare superusers. Higher utilization across every care modality. More likely to track sleep, physical activity, diet, and stress. More engaged with their health data overall. These are not fringe early adopters. These are your most activated, highest-value patients, operating in a channel you do not control and do not measure. 64% of AI health users engage with these tools weekly or more. This is not a one-time experiment. It is a standing habit in a patient population you cannot afford to lose.

The Question Your Team Has Not Answered Yet

When a patient in your market asks ChatGPT for the best orthopedic program, the best cardiology team, or the most trusted cancer center nearby, what does AI say? Is your program cited? Is it described accurately? Does the answer reflect the clinical quality and outcomes data you have invested in building? Most health systems do not know the answer because no one has run the queries. We benchmarked 50 of America's top-ranked programs across five specialties. The national average AI Readiness Score: 6.2 out of 10. Not one program reached 9.0. The institutions your patients trust most are often the ones AI tools mention least.

What to Do This Week

You do not need a six-month initiative to start closing this gap. Three things your team can do now:

1. Run the queries yourself. Open ChatGPT and Perplexity. Ask for the best programs in your top service lines. Document what comes back. That is your baseline.
2. Check your content structure. AI tools favor physician-authored, clinically deep content structured for extraction. Pull your top three condition pages and ask honestly whether they qualify.
3. Measure AI visibility as a channel. If you are not tracking citation rates the way you track Google rankings, you are managing a channel you cannot see. The first step is instrumentation.

The front door moved. The patients are already on the other side of it. The question is whether your strategy has moved with them.

AI Health Strategist offers a complimentary AI Readiness Scan for health systems. We test your top service line across ChatGPT, Google AI Overviews, and Perplexity, score your program across five dimensions, and return a clear picture of where you stand relative to the national benchmark. No cost. No obligation.  

Contact us at jeff.montgomery@aihealthstrategist.com to schedule yours.

‍