Half of Americans didn't turn to large language models for health questions in the past year
Core finding
A slim majority of Americans (51%) say they have not used a large language model such as ChatGPT, Gemini, or Claude for health-related information in the past year. The question went to a randomly assigned half sample, 510 U.S. adults, in a Verasight survey conducted June 25-30, 2025.
Another 20% say they have used one just once or twice in that span, 12% report using one once or twice a month, and 17% report at least weekly use, including 3% who say every day.
Topline distribution
Half of Americans didn't turn to large language models for health questions in the past year.
In the past year, how often have you used a large language model (e.g., ChatGPT, Gemini, Claude) to get health-related information?
n = 510 · Jun 25–30, 2025 · MoE ±5.3%
Methodology
Full methodology →| Mode | Verasight panel recruited via random address-based sampling, random person-to-person text messaging, and dynamic online targeting |
|---|---|
| Population | US adults age 18+ |
| Field dates | Jun 25–30, 2025 |
| Base (unweighted) | 510 |
| Margin of error | ±5.3% |
| Module | ICA Conference Omnibus Survey #2025-058 |
| Sponsor | Verasight |
| Weight variable | weight |
| Weighting targets | age, race/ethnicity, sex, income, education, region, metropolitan status |
Sources
- verasight-ica-2025.tiiny.co
In the past year, how often have you used a large language model (e.g., ChatGPT, Gemini, Claude) to get health-related information?
Cite this question
ICA Conference Omnibus Survey #2025-058, fielded June 25-30, 2025, N=510 US adults age 18+, +/- 5.3%.