The missing disclosure: is generative AI use in bioethics scholarship going largely unreported?

Although much has been written about how generative artificial intelligence (GenAI) may affect bioethics, there is limited research on its use in bioethics scholarship. The objectives of our study are (i) to assess the extent to which GenAI is being used in the production of bioethics scholarship and (ii) to analyze how such utilization is described when disclosed. 20 bioethics journals were selected based on Google Scholar’s 2024 H5-index rankings. All publications from these journals’ 2024 volumes were reviewed for disclosure of GenAI use. In parallel, we examined each journal’s and publisher’s stated policies regarding GenAI-generated content, classifying them as permissive or prohibitive. The frequency and contextual framing of GenAI-utilization acknowledgements were analyzed. Of 1,808 publications, only 12 articles across five journals included a reference to GenAI use or non-use. Among the 12 mentions, nine disclosed GenAI utilization, while three stated non-utilization. The nine utilization acknowledgments appeared in different sections of the papers, with ChatGPT being the most frequently cited GenAI model and “tool” the most commonly used descriptor. Drawing on wider research trends, a disclosure rate of at least 10% was expected; however, the observed rate was under 1% and may reflect limited uptake, underreporting, or both. This finding raises important questions about transparency, methodological reporting standards, and the evolving norms of scholarly practice within the field of bioethics. We recommend clearer standards for when, how, and where GenAI use should be disclosed.