Cross-Sectional Evaluation of Medical Disinformation Safeguards in Consumer-Facing Large Language Model Platforms
Unlabelled: This cross-sectional evaluation of six consumer-facing large language model platforms found significant heterogeneity in safeguard performance against the generation of health disinformation, with Claude and ChatGPT demonstrating complete resistance across all prompt types, while Copilot, Meta AI, Grok, and Gemini exhibited substantial vulnerabilities.
Authors
- Kacper Gradoń (ORCID: https://orcid.org/0000-0003-0750-8678)
- Lisa M. Kalisch Ellett (ORCID: https://orcid.org/0000-0001-5063-6128)
- S. Bacchi
- Andrew Rowland (ORCID: https://orcid.org/0000-0002-8946-3954)
- Natansh D. Modi (ORCID: https://orcid.org/0000-0003-3262-5374)
- Michael D. Wiese (ORCID: https://orcid.org/0000-0002-3255-9242)
- Ross A. McKinnon (ORCID: https://orcid.org/0000-0002-3725-793X)
- Ashley M. Hopkins (ORCID: https://orcid.org/0000-0001-7652-4378)
- Michael J. Sorich (ORCID: https://orcid.org/0000-0003-1999-866X)
- Jessica M. Logan (ORCID: https://orcid.org/0000-0003-4046-6908)
- Bradley D. Menz (ORCID: https://orcid.org/0000-0002-0855-5081)
- Abdulhalim A. Awaty
- Cyril A. Alex
Institutions
- Flinders University (AU)
- University of South Australia (AU)
- The London College (GB)
- University College London (GB)
- The University of Adelaide (AU)
Publication Details
- Journal
- JMIR Infodemiology
- Published
- 2026-04-20
- DOI
- https://doi.org/10.2196/89831
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 0.00