When something doesn’t feel right, Tiffany Davis doesn’t grab her phone to call her doctor. She launches ChatGPT. The 42-year-old from Mesquite, Texas, characterizes it as most people discuss a habit they’ve become accustomed to in a casual, unapologetic manner. She said, “I’ll just basically let ChatGPT know my status, how I’m feeling,” to the AP. “I use it for anything that I’m experiencing.” It’s not particularly noteworthy that she does this. The reason is that about 66 million other Americans follow a similar course of action.
One in four American adults reported using an AI tool or chatbot for health information or advice in the previous 30 days, according to a recent survey by the West Health-Gallup Center on Healthcare in America, which was administered to over 5,500 adults between October and December 2025. Twenty-five percent of the adult population is the kind of statistic that usually makes healthcare administrators sit up straight and start debating its implications right away.
The natural tendency would be to portray this as an indication that Americans are dangerously self-diagnosing or as a crisis of faith in medical professionals. That might be partially accurate. However, what the data actually reveals is more complex and, to be honest, easier to comprehend. Mistrust is not the main reason why people use AI to answer health-related queries. It’s quickness. Seventy-one percent of recent AI health users expressed a desire for prompt responses. Seventy-one percent more requested more details. They are not taking the place of their doctors. The majority of them are adding to the visit by doing research on symptoms prior to entering the examination room or interpreting the doctor’s words after they have left.
AI seems to be bridging a very particular gap here, not the gap between illness and treatment, but rather the gap between ignorance and knowledge. The chief health AI officer at UC San Diego Health, Dr. Karandeep Singh, put it this way: “I almost view it like a better entry portal into web search.” Users receive something more akin to a synthesized summary rather than having to scroll through thirty links in an attempt to reconcile conflicting results. For decades, Americans have been searching for their symptoms on Google. This seems to be the next iteration of that same anxious, restless habit.
However, not everyone who uses AI for health-related queries does so for convenience or curiosity. Approximately 14 million adults, according to Gallup’s estimate, completely avoided seeing a doctor after consulting a chatbot; 14% of recent users claimed they turned to AI because they couldn’t afford a doctor’s visit. It’s more difficult to accept that figure. It raises an issue that the survey doesn’t fully address: whether AI in those circumstances is actually assisting people in making wise decisions or merely providing them with an excuse to put off receiving necessary care. Whether or not those fourteen million people survived it is still unknown.
The data does indicate that most users feel more prepared after using these tools, which is somewhat encouraging. Forty-six percent of respondents reported feeling more confident when speaking with a provider thanks to AI. Twenty-two percent said it made it easier for them to see problems early. That is not insignificant.
However, it’s difficult to ignore the fact that only 4% of AI health users firmly believe the information provided by those tools is accurate. Despite this, people continue to use it. This reveals more about people’s desperation for a starting point—something to hold in place of or prior to the appointment—than it does about AI.
