Maintenance Notice

Due to necessary scheduled maintenance, the JMIR Publications website will be unavailable from Wednesday, July 01, 2020 at 8:00 PM to 10:00 PM EST. We apologize in advance for any inconvenience this may cause you.

Who will be affected?

Accepted for/Published in: JMIR Infodemiology

Date Submitted: Dec 18, 2025
Date Accepted: Mar 28, 2026

The final, peer-reviewed published version of this preprint can be found here:

Cross-Sectional Evaluation of Medical Disinformation Safeguards in Consumer-Facing Large Language Model Platforms

Modi ND, Alex CA, Awaty AA, Menz BD, Bacchi SM, Gradon KT, Logan JM, Rowland A, Kalisch Ellet L, McKinnon RA, Wiese MD, Sorich MJ, Hopkins AM

Cross-Sectional Evaluation of Medical Disinformation Safeguards in Consumer-Facing Large Language Model Platforms

JMIR Infodemiology 2026;6:e89831

DOI: 10.2196/89831

PMID: 42008624

Cross-Sectional Evaluation of Medical Disinformation Safeguards in Consumer-Facing Large Language Model Platforms

  • Natansh D Modi; 
  • Cyril A Alex; 
  • Abdulhalim A Awaty; 
  • Bradley D Menz; 
  • Stephen M Bacchi; 
  • Kacper T Gradon; 
  • Jessica M Logan; 
  • Andrew Rowland; 
  • Lisa Kalisch Ellet; 
  • Ross A McKinnon; 
  • Michael D Wiese; 
  • Michael J Sorich; 
  • Ashley M Hopkins

ABSTRACT

Consumer-facing LLM platforms are increasingly used for health information, but their safeguards can be probed to generate persuasive health disinformation, especially via indirect, narrative-style prompts. In a 90-prompt audit across six topics and six platforms (Nov 27–30, 2025), ChatGPT and Claude produced no disinformation, while Copilot, Meta AI, Grok and Gemini showed substantial vulnerabilities under obfuscation, highlighting the need for continuous evaluation.


 Citation

Please cite as:

Modi ND, Alex CA, Awaty AA, Menz BD, Bacchi SM, Gradon KT, Logan JM, Rowland A, Kalisch Ellet L, McKinnon RA, Wiese MD, Sorich MJ, Hopkins AM

Cross-Sectional Evaluation of Medical Disinformation Safeguards in Consumer-Facing Large Language Model Platforms

JMIR Infodemiology 2026;6:e89831

DOI: 10.2196/89831

PMID: 42008624

Download PDF


Request queued. Please wait while the file is being generated. It may take some time.

© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.