14 Million Americans Skip Doctor Visits Based on AI Health Advice
One in seven recent users of AI for health information say the tool led them to forgo a doctor's visit in the past month, according to a West Health-Gallup survey of 5,660 U.S. adults conducted between October and December 2025. Extrapolated across the population, roughly 14 million adults skipped care based on AI-generated advice.
The finding underscores a tension in how Americans now approach healthcare. While most use AI to supplement traditional care, a meaningful subset is making it a substitute.
How Americans Actually Use AI for Health
One-quarter of U.S. adults have used an AI tool for health information or advice. Among recent users, 59% research before seeing a doctor and 56% research afterward. Both groups treat AI as a preparatory or reflective tool rather than a replacement for professional care.
General conversational systems dominate. Sixty-one percent of users turn to ChatGPT or similar platforms, while 55% use AI embedded in search engines like Google's summaries.
The most common queries involve everyday health concerns: 59% ask about nutrition and exercise, and 58% ask about physical symptoms. Forty-six percent use AI to understand medication side effects, and 44% use it to interpret medical information.
Cost and Access Drive Substitution
Income strongly predicts whether someone uses AI to avoid medical costs. Among households earning under $24,000 annually, 32% have used AI because they couldn't afford a doctor visit. Among those earning $180,000 or more, only 2% cite cost as a reason.
Access barriers and poor provider experiences also matter. Sixteen percent of recent AI users say they turned to the tool because they couldn't reach a provider. Twenty-one percent felt dismissed or ignored by a healthcare provider and sought AI guidance instead.
Trust Remains Split
Confidence in AI-generated health information is evenly distributed. One-third trust it, one-third neither trust nor distrust it, and one-third distrust it. Only 4% strongly trust the accuracy.
Despite this mixed confidence, people act on the information. Eleven percent of recent users say AI recommended something they believed was unsafe.
The Benefit When Used Correctly
When AI complements rather than replaces care, users report concrete gains. Forty-six percent say AI made them more confident asking questions of their provider. Twenty-two percent say it helped identify issues earlier, and 19% say it prevented unnecessary tests.
Younger adults lean more heavily on AI for self-directed research. Sixty-nine percent of users aged 18-29 research before doctor visits, compared with 43% of those 65 and older.
What Healthcare Providers Should Know
Patients arrive with AI-prepared information more often than before. Healthcare systems facing time constraints may find this useful-patients come with questions already researched. But the 14 million skipping care represent a real gap.
The core issue isn't whether AI is good or bad for healthcare. It's that AI is already part of how patients navigate the system, and that reality demands attention to when it helps and when it substitutes for necessary professional judgment.
For healthcare professionals, the trend means patients will increasingly come with AI summaries, interpretations, and concerns. Understanding what tools they used and how much they trust the information becomes part of the intake conversation.
Your membership also unlocks:
