Truace
Subscribe

AI chatbot checks cut people's trust in health myths more than web search did

Study title: Debunking Health Misinformation with Generative AI: The Roles of Perceived Authority and Objectivity

Researchers ran two quasi-experimental pretest-posttest studies. Participants were assigned to verify a false health claim with either a generative AI tool or a search engine. One claim said blood donation is harmful (280 participants). The other said hospice care is equivalent to waiting for death (232 participants). Trust in the claim was measured before and after the task.

Read the study abstract

Health misinformation threatens public health, and generative artificial intelligence (AI) has emerged as a novel tool for correcting it. Its effectiveness and underlying psychological mechanisms, however, remain underexplored. This study compared generative AI-assisted information verification with search-engine-based information verification in reducing public trust in health misinformation and examined the mediating roles of perceived authority and perceived objectivity across two health misinformation topics…

By the Truvace desk1 min read1 sourceEvidence: Moderate · 68/100 — Controlled or validated comparison · n = 512 · 1 independent sourcePeer-reviewedTRV-2026-2265· cite & verify
Debunking Health Misinformation with Generative AI: The Roles of Perceived Authority and Objectivity"Medical laboratories and clinics in Kinwat" by Mouryan, CC BY-SA 4.0.

In brief

The results suggest that how credible people find an information source may help explain why AI-assisted checking worked. The authors describe the effects as short-term and limited to the two claims examined. The findings do not show lasting belief change or results for other health topics.

Main points

  1. Both studies found a significant drop in trust in the false claim in the generative AI condition, compared with the search-engine condition.
  2. Perceived authority and perceived objectivity both significantly mediated the effect.
  3. Perceived authority was more stably associated with trust in the blood donation topic.
  4. Perceived objectivity was more stably associated with trust in the hospice care topic.

The gain

People who used generative AI to check a false health claim ended up trusting it less than people who used a search engine, in studies on blood donation and hospice care.

The rundown

Experiment 1 had 280 participants and tested the claim that blood donation is harmful to health, which concerns alleged physiological risks. Experiment 2 had 232 participants and tested the claim that hospice care is equivalent to waiting for death, which concerns perceptions of hospice care, treatment goals and end-of-life decision-making.

The authors used analysis of covariance, hierarchical regression and bootstrap mediation analyses. Trust was measured before and after verification, while perceived authority and objectivity were measured after the task.

The pattern of regression results differed between topics, so the psychological route to correction may depend on the kind of health claim.

Sources

  1. Peer-reviewedHealth Communication2026-10-23

The debate