Research · 22 June 2026 · Journal of Medical Internet Research
Trusting Generative AI for Health Advice: Preregistered Survey Experiment
Selected by Luis Da Silva
When health information comes through AI, credibility still depends on how people judge both the message and its source. In a preregistered experiment with 1,502 US adults, advice attributed to a human nurse was rated more credible than advice attributed to the two AI sources. For General Value, this supports an editorial question: how should a health organisation make its information understandable and accountable as discovery changes? Clear authorship, visible evidence and careful explanations are proposed applications, not outcomes tested here. The study measures responses to hypothetical scenarios, not search visibility, clinical safety or the commercial performance of a brand.
Study and evidence
Preregistered randomized survey experiment with 1,502 US adults and hypothetical health-advice scenarios.
How to read it
Peer-reviewed publication, June 2026; data collected in July 2025. The study measures perceived credibility in hypothetical scenarios, not clinical decisions, safety, search rankings, brand conversion or longevity outcomes. Greater trust is not proof of greater accuracy.
Why it matters
Our editorial interpretation: make health information understandable and accountable as discovery tools change. Visible evidence and clear responsibility are communication principles to test, not effects on AI search or clinical outcomes established by this study.
Luis Da Silva, Partner & Chief Marketing Officer
About the publisher
Asheley R. Landrum, Nitin Verma, Amanda Kehrberg. Journal of Medical Internet Research.
Useful for
Communication, audience-research and service-design teams. The proposed application is General Value's interpretation and needs local validation.
Credit and license Asheley R. Landrum, Nitin Verma, Amanda Kehrberg. 2026. Trusting Generative AI for Health Advice: Preregistered Survey Experiment. Journal of Medical Internet Research, 28, e97882. DOI: 10.2196/97882. Copyright: Asheley R. Landrum, Nitin Verma and Amanda Kehrberg, 2026. CC BY 4.0, https://creativecommons.org/licenses/by/4.0/. Original publication: https://www.jmir.org/2026/1/e97882/. General Value supplies an original editorial summary; the full paper opens at its publisher.
Original research belongs to the credited authors. The bilingual summary and proposed application are General Value's editorial interpretation, not findings tested for our firm.