Americans only somewhat trust AI health advice

Aug 13–18, 2025 · n = 1,000 · MoE ±3.1%Source report ↗

Core finding

Americans keep their guard up when AI offers health guidance. Just 14% of U.S. adults place high trust in recommendations from large language models such as ChatGPT, Gemini and Grok for lifestyle changes to improve their health, including 4% who say they trust them extremely, according to an August 2025 Verasight survey.

Cooler responses dominate: 32% trust these tools moderately, 26% only slightly and 25% not at all, while 3% prefer not to answer.

Topline distribution

Americans only somewhat trust AI health advice.

How much do you trust AI (Large Language Models such as ChatGPT, Gemini, and Grok) when it comes to its recommendations for lifestyle changes to improve your overall health?

Moderately
31.9%
Slightly
26.5%
Not at all
24.7%
Very
10.0%
Extremely
4.4%
Prefer not to answer
2.5%

n = 1,000 · Aug 13–18, 2025 · MoE ±3.1%

ModeVerasight panel recruited via random address-based sampling, random person-to-person text messaging, and dynamic online targeting
PopulationUnited States adults
Field datesAug 13–18, 2025
Base (unweighted)1,000
Margin of error±3.1%
ModuleVerasight JSM Omnibus Survey
SponsorVerasight
Weight variableweight
Weighting targetsage, race/ethnicity, sex, income, education, region, metropolitan status

Sources

  1. 01

    How much do you trust AI (Large Language Models such as ChatGPT, Gemini, and Grok) when it comes to its recommendations for lifestyle changes to improve your overall health?

    reports.verasight.io/r/jsm-2025

Cite this question

Verasight JSM Omnibus Survey, fielded August 13-18, 2025, N=1,000 United States adults, +/- 3.1%.