Are AI chatbots exposing healthcare's patient engagement limits?

Patients can't access needed care. ChatGPT stepped in.

Healthcare's patient access problems aren't exactly new.

Indeed, data from the National Association of Community Health Centers shows that 100 million people lack a usual source of care, meaning they don't have a provider to turn to with a medical need.

The forces behind that lapse are manifold. Healthcare is expensive, and many consumers are choosing to go without care to protect their finances. There's also a growing provider shortage, making it even harder for patients who do want care to book an appointment.

And that's not to mention the appointment booking process itself. Both call centers and online appointment scheduling systems have their quirks. When patients do manage to navigate those systems, they still face an average 31-day wait time to see a provider.

Other social determinants of health, particularly transportation barriers, also keep patients from seeing their providers.

"It's remarkable that in today's age, calling and waiting to talk to a doctor is such a tedious task that you have to carve out 30 minutes of your day for it," Germain Farraher said.

Patients are tired of navigating a complex health system just to get their basic healthcare needs met, and so they find themselves accessing a new tool: ChatGPT.

At the start of 2026, OpenAI released a report about its flagship product, showing that users send 2 million messages to ChatGPT focused on healthcare.

A separate survey from Sacred Heart University showed that patients are using these tools because the current slate of patient engagement resources are falling short. About a third of patients are already using these tools to help them research medical topics, and the survey showed they're willing to use them even more for care coordination.

In other words, there's a market for these types of AI chatbots.

"What utilizing ChatGPT, or any AI, shows is that there's a broken healthcare system," according to Nicole Lamoureaux, the CEO of the National Association of Free and Charitable Clinics.

"It's noteworthy that so many people have to utilize AI for healthcare instead of going to the doctor," added Lamoureaux, who said she worries AI's insurgence mostly underscores a health system failing.

These trends could continue, according to Lamoureaux. With Affordable Care Act Marketplace subsidies lapsed and changes in Medicaid eligibility coming down the pike, there's likely to be an increase in the uninsured population. That means more patients who can't afford to connect with a doctor conveniently and affordably, opening the door for more AI use.

However, care access challenges span beyond affordability. According to Germain Farraher, the healthcare industry has failed to set up a usable ecosystem of healthcare resources, leaving patients mostly in the dark about their health and, in some cases, sowing mistrust.

"There are so many things alluding trust," she asserted. "Websites are not that great. When you go to your healthcare provider websites, sometimes you might run into resources that are pertinent and relevant to you, but a lot of times you might just go down a rabbit hole searching for something and realizing they don't have anything that's relevant to you. Why would I come here for information?"

To be clear, patient trust isn't entirely gone. Patients largely trust their direct healthcare provider. But when it's so much work just to see that provider, why wouldn't a patient try ChatGPT or Claude or Gemini first?

Coupled with the fact that AI chatbots help alleviate patient health literacy shortcomings, it's clear they are here to stay. The question is: how can providers respond?

AI chatbots promise to address health literacy. Is the industry ready?

It's not just that AI chatbots give medical advice without the friction of booking a doctor's appointment.

According to Foluke Omosun, Ph.D., one of the lead researchers on the Sacred Heart study and an assistant professor of communication and media studies at the University, AI chatbots can help patients understand complex health information.

That's what makes them so promising.

"People have at some time or another experienced difficulty understanding a doctor's instructions," she explained during an interview. "AI can aid healthcare delivery because now you can be more engaged with your health. Patients can use these tools to help them understand instructions, understand their symptoms and help them understand what questions to ask their doctors."

Per figures from the Department of Health and Human Services, nearly a third of adult patients have basic or below basic health literacy levels, while only 12% have proficient health literacy.

So even when a patient can and does meet with a healthcare provider, there's a strong possibility they don't understand or don't remember what was discussed. The likes of ChatGPT Healthcare, Claude for Health and Amazon One Medical's Health AI assistant hope to clear that up by taking in patients' medical records and giving them plain-language explanations.

That has the potential to support patient activation and autonomy, according to Omosun.

"AI tools are there to help fill that gap and help patients be more empowered when they feel powerless in the healthcare system," she said.

Even Germain Farraher, who has expressed more trepidation about consumer-facing AI in healthcare, acknowledged the potential for AI to make medical terminology more accessible.

"A third of clinicians think that there is low patient health literacy, meaning that a third of patients don't have a great understanding of medical language and treatment and diagnostics. Nor should they -- they don't have a medical background," Germain Farraher said, citing internal Forrester numbers.

But there's a flip side to that potential. Although AI chatbots seek to provide and explain medical information, that doesn't mean the tools always do so accurately. These systems are liable to bias and misinformation, and it will require a great deal of not just health literacy, but also media literacy, for patients to parse through it all.

"But then to present this AI-generated information to them, should we then rely on them to be able to read between the lines here?" Germain Farraher posited. "Is it now their responsibility to know what a reliable source is that's linked at the bottom? Is it up to them to now click on that and go in and learn more?"

In an ideal world, patients would know how to dig into an AI's sourcing and use it mostly for background research, not in place of a doctor. But Germain Farraher worries that ideal world is unlikely.

With the reality that consumer-facing health AI is not going away, it'll be incumbent upon healthcare providers to support patients in responsible use.

Orgs gear up for more health AI

AI chatbots are stepping in where traditional healthcare systems have fallen short, regardless of the risks associated with the technologies.

Still, according to Lamoureaux, the healthcare industry needs to pull a few policy levers to ease patient care access -- not just rely on AI to fill in the gap. Fixing things like Medicaid eligibility and ACA plan affordability and increasing investment in social determinants of health will help get us there, she said.

But it'd still be a mistake to completely ignore AI chatbots for patients.

"If you do, your organization becomes out of touch," Lamoureaux said. "The toothpaste is out of the tube already."

Indeed, healthcare organizations should look at AI chatbots as a part of the overarching solution to its broken patient engagement systems, Lamoureaux advised. Keeping in line with the omnichannel approach to healthcare consumerism many in the industry have been embracing for years, AI chatbots need to be one piece of a multi-part fix.

"Of course, a computer is not going to replace a human, and a computer is not going to replace a doctor," she said. "But at the same time, if organizations don't start looking at these, you could lose trust with your patients on the other side as well."

To get there, healthcare leaders need to assess the risks and benefits of the tools and design a game plan for integrating them into clinical practice that accounts for both.

Sara Heath has reported news related to patient engagement and health equity since 2015.