Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 19 of the 19 references it cites were found in a registry or an index.

Current safeguards, risk mitigation, and transparency measures of large language models against the generation of health disinformation: repeated cross sectional analysis

Open access

Abstract

Objectives: To evaluate the effectiveness of safeguards to prevent large language models (LLMs) from being misused to generate health disinformation, and to evaluate the transparency of artificial intelligence (AI) developers regarding their risk mitigation processes against observed vulnerabilities. Design: Repeated cross sectional analysis. Setting: Publicly accessible LLMs. Methods: In a repeated cross sectional analysis, four LLMs (via chatbots/assistant interfaces) were evaluated: OpenAI’s GPT-4 (via ChatGPT and Microsoft’s Copilot), Google’s PaLM 2 and newly released Gemini Pro (via Bard), Anthropic’s Claude 2 (via Poe), and Meta’s Llama 2 (via HuggingChat). In September 2023, these LLMs were prompted to generate health disinformation on two topics: sunscreen as a cause of skin cancer and the alkaline diet as a cancer cure. Jailbreaking techniques (ie, attempts to bypass safeguards) were evaluated if required. For LLMs with observed safeguarding vulnerabilities, the processes for reporting outputs of concern were audited. 12 weeks after initial investigations, the disinformation generation capabilities of the LLMs were re-evaluated to assess any subsequent improvements in safeguards. Main outcome measures: The main outcome measures were whether safeguards prevented the generation of health disinformation, and the transparency of risk mitigation processes against health disinformation. Results: Claude 2 (via Poe) declined 130 prompts submitted across the two study timepoints requesting the generation of content claiming that sunscreen causes skin cancer or that the alkaline diet is a cure for cancer, even with jailbreaking attempts. GPT-4 (via Copilot) initially refused to generate health disinformation, even with jailbreaking attempts—although this was not the case at 12 weeks. In contrast, GPT-4 (via ChatGPT), PaLM 2/Gemini Pro (via Bard), and Llama 2 (via HuggingChat) consistently generated health disinformation blogs. In September 2023 evaluations, these LLMs facilitated the generation of 113 unique cancer disinformation blogs, totalling more than 40 000 words, without requiring jailbreaking attempts. The refusal rate across the evaluation timepoints for these LLMs was only 5% (7 of 150), and as prompted the LLM generated blogs incorporated attention grabbing titles, authentic looking (fake or fictional) references, fabricated testimonials from patients and clinicians, and they targeted diverse demographic groups. Although each LLM evaluated had mechanisms to report observed outputs of concern, the developers did not respond when observations of vulnerabilities were reported. Conclusions: This study found that although effective safeguards are feasible to prevent LLMs from being misused to generate health disinformation, they were inconsistently implemented. Furthermore, effective processes for reporting safeguard problems were lacking. Enhanced regulation, transparency, and routine auditing are required to help prevent LLMs from contributing to the mass generation of health disinformation.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1136/bmj-2023-078538

Cites

19 of 19 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 Flowchart of observed capabilities of large language models to facilitate the generation of disinformation on cancer from primary analyses conducted September 2023. LLMs=large language models
Fig 1 Flowchart of observed capabilities of large language models to facilitate the generation of disinformation on cancer from primary analyses conducted September 2023. LLMs=large language modelsCC BY-NC · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

4 more claims about this paper

Create a free account to see every claim other papers make about this one.

See all claims — free account

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 19 references

Research works printed but not linked: 1

  1. Liu Y Deng G Xu Z . Jailbreaking chatgpt via prompt engineering: An empirical study. arXiv 230513860 2023

Cited by

See all 134

Metadata

Authors
Menz, B. D., Kuderer, N. M., Bacchi, S., Modi, N. D., Chin-Yee, B., Hu, T., Rickard, C., Haseloff, M., Vitry, A., McKinnon, R. A., Kichenadasse, G., Rowland, A., Sorich, M. J., & Hopkins, A. M.
Affiliations
  • Flinders University (AU)
  • Advanced Cancer Research Group, Kirkland, WA, USA
  • Western University (CA)
  • University of Cambridge (GB)
Corresponding author
Hopkins, Ashley M
Venue
BMJ
Publisher
BMJ
Published
2024-03-20
Volume
384
Pages
e078538
ISSN
1756-1833 (ISSN-L)
OpenAlex
W4393054030
Cited by
132
References
19
Citations by year
2026: 60 · 2025: 55 · 2024: 17
FWCI
94.68
Type
journal article
Language
en
Open access
Locations
  • BMJ (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • Apollo (University of Cambridge) (submittedVersion, cc-by-nc)
  • Adelaide Research & Scholarship (AR&S) (University of Adelaide) (submittedVersion, cc-by-nc)
  • HighWire Press Open Archive (submittedVersion)
Funders
  • Cancer Council Australia
  • National Health and Medical Research Council
Topics
  • Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
  • Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
Keywords
large language models, health disinformation, jailbreaking, risk mitigation, AI transparency, alkaline diet, AI-generated content, LLM vulnerabilities, GPT-4, Claude 2, PaLM 2, Llama 2, Gemini Pro, ChatGPT, Copilot, bards
MeSH terms
Disinformation, Animals, Artificial Intelligence, Camelids, New World, Cross-Sectional Studies, Humans, Language, Skin Neoplasms, Sunscreening Agents

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Current safeguards, risk mitigation, and transparency measures of large language models against the generation of health disinformation: repeated cross sectional analysis. https://app.cochat.ai/papers/10.1136/bmj-2023-078538
The paper
Menz, B. D., Kuderer, N. M., Bacchi, S., Modi, N. D., Chin-Yee, B., Hu, T., Rickard, C., Haseloff, M., Vitry, A., McKinnon, R. A., Kichenadasse, G., Rowland, A., Sorich, M. J., & Hopkins, A. M. (2024). Current safeguards, risk mitigation, and transparency measures of large language models against the generation of health disinformation: repeated cross sectional analysis. BMJ. https://doi.org/10.1136/bmj-2023-078538

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.