A Third of Americans Now Ask AI About Their Health

A growing number of Americans are taking their health questions to a chatbot. A Pew Research Center survey of 3,488 US adults, published on 25 August, found that 34% have used an AI chatbot for at least one health-related task. The most common use was looking up quick health information, cited by 28%, followed by trying to understand what was causing a symptom, at 25%, and finding low-cost information, at 22%.

The appeal is easy to understand. A chatbot answers at midnight, does not charge a copay and does not make you feel rushed. Among those who had tried it, roughly 47% called the answers extremely or very helpful. For a lot of people, the alternative was a search engine and a spiral of worst-case results, and a plain-language explanation is a genuine improvement on that.

Helpful is not the same as trusted

The comfort runs shallower than the usage, though. Only 29% of respondents said they were very comfortable sharing personal health information with these tools. People are happy to ask a chatbot what a symptom might mean, and much warier about handing over the details that would make the answer accurate. That gap, between finding a tool useful and trusting it with your data, is the real story in the numbers.

Adoption is far from even. Asian Americans reported the highest use at 56%, and adults under 30 were close behind at 44%. Older adults were the most likely to say they simply were not sure what to make of the technology.

The professional caveat

Doctors have watched this shift with a mix of relief and worry. A well-built model can translate jargon and flag when something warrants a visit. It can also state a wrong answer with total confidence, and a survey published the same day found Americans broadly skeptical that chatbots help people dealing with loneliness or depression. The clinical evidence base is thin, too. Research this month found that most FDA-cleared medical AI has never been tested on actual patient outcomes.

None of that has slowed the public. A third of the country is already using these tools for health, and the reservation they hold is a sensible one: interested in the answers, cautious about the data. As the Pew findings and the wider debate over trust in AI make clear, that caution is doing useful work.