Millions Of Americans Are Skipping Doctor’s Visits After Acting On AI Health Advice

You feel a sharp pain in your side at 11 p.m. on a Tuesday. Instead of calling a nurse hotline or waiting for a morning appointment, you open ChatGPT and type in your symptoms. For most Americans who do this, it’s a way to gather information before seeing a real doctor. But for an estimated 14 million adults, based on survey projections, the chatbot’s answer was enough to skip the doctor altogether, even though few users say they strongly trust what AI tells them.

A new survey from the West Health-Gallup Center on Healthcare in America finds that one in four American adults has used an AI tool or chatbot to get health information or advice. The survey polled more than 5,500 adults across the country between late October and late December of 2025, and it captures how artificial intelligence is quietly reshaping everyday healthcare decisions, sometimes helpfully, sometimes in ways that raise concerns.

Most people using AI for health questions aren’t trying to replace their doctor. They’re preparing. Fifty-nine percent of recent users said they turned to AI to research on their own before a doctor visit, and 56 percent said they used it to dig deeper after one. But within those reassuring numbers is a more troubling trend: 14 percent of recent users said AI-generated advice led them to skip a provider visit entirely in the past 30 days.

Projected across the population, that translates to an estimated 14 million adults who reported deciding against seeing a healthcare provider because of something a chatbot said. Trust in AI health advice, meanwhile, is split almost evenly, with roughly a third of users trusting it, a third neutral, and a third distrusting it, but strong confidence is rare. Only 4 percent say they strongly trust its accuracy.

What Americans Are Asking AI Chatbots About Their Health

The biggest drivers of AI health use are speed and curiosity. Among people who recently used AI for health questions, 71 percent said they wanted answers quickly, and an identical 71 percent said they wanted additional information. Sixty-seven percent were simply curious about what AI would say. The most common topics were nutrition or exercise (59 percent) and physical symptoms (58 percent). Close behind were medication side effects (46 percent), making sense of medical information like test results or doctor’s notes (44 percent), and researching a diagnosis or medical condition (38 percent). About a quarter of users asked AI about mental health or emotional concerns.

The tools people are reaching for are the ones already on their phones and laptops. Sixty-one percent of users said they relied on general conversational AI systems like ChatGPT or Copilot, while 55 percent used AI features built into web searches, such as Google’s AI-generated summaries.

Nearly half of all users, 46 percent, said the AI tool made them feel more confident when talking with a healthcare provider. Twenty-two percent said it helped them catch potential issues earlier, and 19 percent said it helped them avoid unnecessary medical tests or procedures. On the surface, that sounds like a win for an overburdened healthcare system.

Why Cost and Access Push People Toward an AI Chatbot

Not everyone turns to AI out of curiosity. For some, it fills a gap because the real healthcare system feels out of reach. The survey found that 27 percent of recent users said they turned to AI because they didn’t want to pay for a doctor’s visit, and 14 percent said they couldn’t afford a doctor’s visit. Sixteen percent said they couldn’t access a provider at all, and 42 percent wanted help outside normal business hours. Twenty-one percent said they didn’t have time to make an appointment.

Income plays a dramatic role. Among recent users living in households earning less than $24,000 a year, 32 percent said they used AI because they were unable to pay for a doctor’s visit. Among those in households earning $180,000 or more, that figure dropped to just 2 percent. (The survey notes that the sample size for the lowest income group was below 300 respondents, so those figures should be interpreted with some caution.) Still, the gap points to a hard reality: for some lower-income Americans, AI may be acting as a stopgap when care is out of reach.

Age matters too, though the pattern looks different. Younger adults are far more likely to use AI for research before appointments. Sixty-nine percent of recent users between 18 and 29 said they use AI to research before seeing a doctor, compared with 43 percent of those 65 and older. Still, the fact that more than four in ten older adults are doing the same thing signals that this behavior has spread well beyond younger generations.

Why Millions Use AI Health Advice Without Strong Confidence

One of the more eye-opening findings is how split Americans are on trusting the very tool they’re using to make healthcare decisions. Among recent users, the responses break into almost perfect thirds: 33 percent said they trust AI health information, 33 percent said they neither trust nor distrust it, and 34 percent actively distrust it. Strong trust is rare, with only 4 percent saying they strongly trust the accuracy of AI health information.

That means millions of Americans are acting on health advice from a source most view with only moderate confidence at best. And the consequences aren’t purely hypothetical. About 11 percent of users said AI recommended information or advice that they believed was unsafe. That’s roughly one in ten people encountering guidance they personally flagged as potentially dangerous.

Another 21 percent of users said they turned to AI because they had felt dismissed or ignored by a provider in the past, and 18 percent said they were too embarrassed to talk to a real person about their health concern. These are people who aren’t just looking for faster answers. They’re looking for answers that come without judgment, even if those answers might be wrong.

What This Means for AI and the Future of Healthcare

AI is already shaping how patients prepare for visits, how they interpret what a doctor tells them, and how they decide whether to seek care in the first place. For many users, these tools seem to be genuinely productive, helping people arrive at appointments better informed or making sense of confusing medical documents afterward.

But for a meaningful slice of the population, AI is doing something far more serious: it’s replacing the provider visit entirely. An estimated 14 million adults reported not seeing a provider they otherwise would have seen, based on what a chatbot told them. Whether those decisions turned out fine or led to missed diagnoses and delayed treatment is something this survey wasn’t designed to measure. It captures what people say they did, not what happened to them afterward. But it’s a question the healthcare system will have to reckon with as these tools become a bigger part of daily life.

A quarter of American adults are already using AI for health advice. Few of them express strong confidence in it. And millions are letting it determine whether they need to see a doctor at all. The technology isn’t waiting for the healthcare system to catch up.

---

Disclaimer: This article is for informational purposes only and does not provide medical advice, diagnosis, or treatment. The survey findings cited reflect self-reported responses and population estimates, not verified medical outcomes. AI tools can generate inaccurate, incomplete, or unsafe health information and should not be used as a substitute for a qualified healthcare professional. Anyone experiencing concerning, severe, or urgent symptoms should seek prompt medical care or emergency assistance.

---