Artificial intelligence policies in bioethics and health humanities: a comparative analysis of publishers and journals

Abstract

Introduction

Rapid advancements in artificial intelligence (AI) pose novel ethical and practical challenges for scholarly publishing. Although AI-related policies are emerging in many disciplines, little is known about the extent and clarity of AI guidance in bioethics and health humanities journals.

Methods

A search of publicly available journal lists from the American Society for Bioethics and Humanities, Health Humanities Consortium, and Association for Medical Humanities was supplemented with Google Scholar’s top 20 bioethics journals ranked by h5-index. This yielded 54 unique journals, of which 50 remained after excluding those without a functional website or recent publications. AI policies were reviewed at the journal and publisher levels were assessed via website review, and editors were contacted for clarification when required. Data extraction was conducted by one author and independently verified by two additional researchers to ensure accuracy.

Results

Of the 50 journals analyzed, only 8 (16%) had a clear AI policy, while 27 (54%) were published by a publisher with an identifiable AI policy. Publisher AI policy statements were favorable to considering AI-assisted manuscripts. Five (10%) of the 8 journals with a clear AI policy explicitly prohibited AI-generated text in submissions. The remaining 15 (30%) journals did not have a publicly available AI policy. Ten of these 15 journals confirmed an absence of any formal AI policy, and seven indicated that discussion to develop guidelines was ongoing.

Conclusion

The adoption of AI policies in bioethics and health humanities journals is currently inconsistent. Some journals explicitly ban AI-generated text, whereas others permit AI-assisted writing, with publisher policies being favorable to considering AI-assisted manuscripts. The lack of standardized AI guidelines underscores the need for further discussion to ensure the ethical and responsible integration of AI in academic publishing.

Introduction

When ChatGPT was released in November 2022, many publishers and journals were unprepared to manage the use of artificial intelligence (AI) in the academic research and writing process. Early on this was seen by some researchers recording ChatGPT as a co-author [6, 30]. This was quickly identified as a problem, leading publishers and journals to publish guidance that ruled out crediting an AI model as a co-author because it cannot meet established authorship criteria [28]. For example, the International Committee of Medical Journal Editors (ICMJE) has since stated that an AI model cannot be considered responsible for the accuracy, originality, or integrity of the work and, therefore, cannot be considered an author in a meaningful sense [15]. So, while this particular problem was swiftly identified and addressed, in the years since, many publishers and journals still have differing guidance and expectations about how AI can or cannot be used in the scholarly research [9], and yet scholars are increasingly utilizing AI in the research and writing process [26, 27].

In December of 2023, a group of editors of bioethics and health/medical humanities journals published a joint statement on the use of AI in scholarly writing [19]. Among other points of consensus, the editors agreed that AI should not be a co-author, that authors should be transparent in their use of AI, and that authors are responsible for the content of the paper. Two issues are noteworthy, however. First, a review of some of the journal homepages of the editors who signed the joint statement reveals a lack of guidance for authors on AI use. For example, potential authors are hard pressed to find an AI policy on the journal homepage for Bioethics or AMA Journal of Ethics, among others. It is not difficult to see that the absence of an AI policy can have implications for researchers and journals. For instance, if a journal lacks an AI policy, a researcher may interpret this as indicating that there is no need to declare their use of AI, or different editors may handle manuscripts that utilize AI in the writing process inconsistently. Second, while the joint statement suggests that it is permissible for authors to responsibly use AI in scholarly writing, at least one editor who signed on to the joint statement oversees two journals—Journal of Medicine and Philosophy and Christian Bioethics—that expressly forbid any AI-assisted submissions. It is important, therefore, to assess AI policies among bioethics and health/medical humanities journals.

To the best of our knowledge, journal policies on AI utilization in academic bioethics and health/medical humanities scholarship remain unexplored. This study compares the policies of leading bioethics and health humanities journals regarding AI-generated content and explores the impact that variation in AI policies can have.

Methods

Bioethics is an interdisciplinary field that examines ethical, legal, and social issues arising in healthcare, medical research, and the life sciences (e.g., [4, 38]). Bioethicists engage in ethical examination of topics such as patient rights, decision-making, biotechnology, public health ethics, and the ethical implications of emerging technologies. Similarly, health or medical humanities, which is a relatively novel discipline that is only 25 years or so old, is an interdisciplinary field that explores the human experience of health, illness, and medicine through the lens of the humanities, including literature, philosophy, history, visual arts, and cultural studies (e.g., [8, 21]). Health/medical humanists seek to enhance our collective understanding of patient experiences, medical practice, and healthcare systems by incorporating narrative, ethical reflection, and artistic expression [12]. There is overlapping similarity but also differences between the disciplines, which is why we have opted to include them together in this study.

Journal selection

On the 8th of March 2025 a manual search of publicly available lists of bioethics and health/medical humanities journals from the American Society for Bioethics and Humanities [2], Health Humanities Consortium [11], and Association for Medical Humanities [3] was conducted. Because these lists primarily focused on bioethics and humanities or just health/medical humanities, we supplemented our search by using Google Scholars top 20 bioethics journals ranked according to their h5-index; this led to the inclusion of an additional 9 journals. To help prevent the inclusion of journals from predatory publishers, each publisher was checked for its inclusion against recognised criteria [29]. A predatory journal is one that prioritises profit over scholarly rigor by bypassing transparent editorial and peer-review processes [1]; no journal was removed based on this criteria.

After removing duplicates, 54 unique journals were identified. Journals without a functioning webpage or that did not have a volume published within the last three years were excluded, resulting in a final sample of 50 journals. Each journal was classified as either primarily bioethics or primarily health humanities based on the aims and scope of the journal. This resulted in a total of 36 bioethics and 14 health humanities journals (see Table 1).

Table 1.

List of journals and their publishers included in the study, along with their classification

	Journal (in alphabetical order)	Publisher	Bioethics (B) or Health Humanities (HH)
1	AMA Journal of Ethics	American Medical Association	B
2	American Journal of Bioethics	Taylor & Francis	B
3	AJOB Empirical Bioethics	Taylor & Francis	B
4	AJOB Neuroscience	Taylor & Francis	B
5	Ars Medica	University of Toronto Press	HH
6	Asian Bioethics Review	Springer Nature	B
7	Bioethics	Wiley	B
8	BMC Medical Ethics	BMC	B
9	Cambridge Quarterly of Healthcare Ethics	Cambridge University Press	B
10	Canadian Journal of Bioethics	University of Montreal	B
11	Christian Bioethics	Oxford University Press	B
12	Clinical Ethics	Sage	B
13	Developing World Bioethics	Wiley	B
14	Ethics, Medicine and Public Health	Elsevier	B
15	Global Bioethics	Taylor & Francis	B
16	Hastings Center Report	Wiley	B
17	Healthcare Analysis	Springer Nature	B
18	HEC Forum	Springer Nature	B
19	Hektoen International A Journal of Medical Humanities	Hektoen Institute of Medicine	HH
20	International Journal of Feminist Approaches to Bioethics	University of Toronto Press	B
21	Journal of Bioethical Inquiry	Springer Nature	B
22	Journal of Empirical Research on Human Research Ethics	Sage	B
23	Journal of Humanities in Rehabilitation	Emory University	HH
24	Journal of Law, Medicine, and Ethics	Cambridge University Press	B
25	Journal of Literary & Cultural Disability Studies	Liverpool University Press	HH
26	Journal of Medical Ethics	BMJ	B
27	Journal of Medical Humanities	Springer Nature	HH
28	Journal of Medicine and Philosophy	Oxford University Press	B
29	Journal of Pediatric Ethics	Children’s Minnesota	B
30	Kennedy Institute of Ethics Journal	Johns Hopkins University Press	B
31	Literature and Medicine	Johns Hopkins University Press	HH
32	Medical Humanities	BMJ	HH
33	Medicine, Health Care and Philosophy	Springer Nature	B
34	Neuroethics	Springer Nature	B
35	Nursing Ethics	Sage	B
36	Pediatric Ethicscope	No publisher identified	B
37	Pegasus Review: A Medical Literary Review Journal	No publisher identified	HH
38	Perspectives in Biology and Medicine	Johns Hopkins University Press	B
39	Public Health Ethics	Oxford University Press	B
40	Pulse- Voices from the Heart of Medicine	Voices from the Heart of Medicine	HH
41	Research & Humanities in Medical Education (RHiME)	Medical Humanities Group	HH
42	Rhetoric of Health & Medicine	University of Florida Press	HH
43	Synapsis: A Health Humanities Journal	Columbia University Libraries	HH
44	The Healing Muse	Center for Bioethics and Humanities, Upstate Medical University	HH
45	The Postgraduate Journal of Medical Humanities	University of Exeter	HH
46	The Journal of Clinical Ethics	University of Chicago Press	B
47	The Journal of Hospital Ethics	MedStar Health	B
48	Theoretical Medicine & Bioethics	Springer Nature	B
49	Narrative Inquiry in Bioethics	Johns Hopkins University Press	B
50	Voices in Bioethics	Columbia University	B

Search strategy

A comprehensive search of bioethics and health/medical humanities journal AI policies required a thorough examination of both individual journals and their affiliated publishers. Journals are typically owned, managed, or affiliated with publishers who oversee their production, distribution, and financial sustainability. While journals maintain some editorial autonomy, their policies may be constrained by broader publisher guidelines (e.g., [18]). This means that even if a journal does not explicitly state an AI policy on its homepage, it may still adhere to a publisher-level policy.

One reviewer (CB) examined each journal’s homepage, author guidelines, editorial statements, and submission policies for references to AI use. If no AI-related statement was found at the journal level, the reviewer categorized the journal as “no” for “AI Journal Policy.” If a statement could be found on the journal homepage, the reviewer recorded “yes” to “AI Journal Policy” and then copied the relevant language to the Excel spreadsheet “Journal policy wording” category. If no statement could be found on the journal homepage, the reviewer recorded “no” to “AI Journal Policy” and then searched for the journal publisher’s AI policy statement. If a publisher policy was found, the reviewer recorded “yes” under “AI Publisher Policy” and documented the relevant wording to the Excel spreadsheet “Publisher policy wording” category. If no policy was identified, the journal was marked as “no” for both journal- and publisher-level AI policies. If a journal was found to have a journal or publisher AI policy, the reviewer summarized the policy as either permissive or non-permissive: a policy is permissive if the journal is favorable to considering AI-assisted manuscripts, while a policy is non-permissive if the journal does not accept submissions that contain AI-generated text.

To ensure the reliability of the search, a second and third reviewer (DR; DH) independently verified all journal and publisher assessments and summaries. Any discrepancies were discussed and resolved collaboratively and ensured a thorough and transparent review of any AI policies. In addition, reviewers emailed the journals or the editors of journals that were marked as “no” for both journal- and publisher-level AI policies to confirm whether they have an available AI policy.

Results

Of the 50 journals analyzed, only 8 (16%) journals have an identifiable AI policy statement, while 27 (54%) journals are published by a publisher with a clear AI policy statement. The remaining 15 (30%) journals and their publishers did not have an identifiable AI policy statement publicly available. Only 1 of the journals with a clear AI policy statement (Research & Humanities in Medical Education (RHiME)) is primarily a health/medical humanities journal while the remaining 7 are primarily bioethics journals (Christian Bioethics; Ethics, Medicine, and Public Health; Hastings Center Report; Journal of Medicine and Philosophy; Kennedy Institute of Ethics Journal; The Journal of Hospital Ethics; Voices in Bioethics). Of the 27 journals without a clearly stated AI journal policy that were published by a publisher with a clear AI policy statement, 23 were primarily bioethics while the remaining 4 were health/medical humanities. Among those journals for which an AI policy statement could not be found, 6 were primarily bioethics and 9 were primarily health/medical humanities (See Fig. 1).

Fig. 1.

Bar chart comparing bioethics and health/medical humanities journals AI policies

Additionally, 5 (10% total) of the 8 journals that offer an AI policy explicitly state that they do not accept submissions that contain AI-generated text. These are primarily bioethics journals: Christian Bioethics; Journal of Medicine and Philosophy; The Journal of Hospital Ethics; Kennedy Institute of Ethics Journal; Voices in Bioethics. The remaining 2 (4%) journals with a clear AI policy statement indicated a willingness to accept manuscripts from authors who utilized AI. The 27 (54%) publisher AI policy statements are favorable to considering AI-assisted manuscripts. See Table 2 for results.

Table 2.

Breakdown of journal, AI policy or absence, and whether AI policy is permissive or non-permissive

Journal	Journal, publisher, or no identifiable AI policy & link	Permissive or non-permissive	Select policy language
AMA Journal of Ethics (B)	No identifiable policy	–-	–-
American Journal of Bioethics (B)	Publisher (https://taylorandfrancis.com/our-policies/ai-policy/)	Permissive	“Authors must clearly acknowledge within the article or book any use of Generative AI tools through a statement which includes: the full name of the tool used (with version number), how it was used, and the reason for use. For article submissions, this statement must be included in the Methods or Acknowledgments section.”
AJOB Empirical Bioethics (B)	Publisher (https://taylorandfrancis.com/our-policies/ai-policy/)	Permissive	“Authors must clearly acknowledge within the article or book any use of Generative AI tools through a statement which includes: the full name of the tool used (with version number), how it was used, and the reason for use. For article submissions, this statement must be included in the Methods or Acknowledgments section.”
AJOB Neuroscience (B)	Publisher (https://taylorandfrancis.com/our-policies/ai-policy/)	Permissive	“Authors must clearly acknowledge within the article or book any use of Generative AI tools through a statement which includes: the full name of the tool used (with version number), how it was used, and the reason for use. For article submissions, this statement must be included in the Methods or Acknowledgments section.”
Ars Medica (HH)	No identifiable policy	–-	––
Asian Bioethics Review (B)	Publisher (https://www.springer.com/gp/editorial-policies/artificial-intelligence--ai-/25428500)	Permissive	“Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared.”
Bioethics (B)	Publisher (https://authorservices.wiley.com/ethics-guidelines/index.html#22)	Permissive	“If an author has used a GenAI tool to develop any portion of a manuscript, its use must be described, transparently and in detail, in the Methods section (or via a disclosure or within the Acknowledgements section, as applicable). The author is fully responsible for the accuracy of any information provided by the tool and for correctly referencing any supporting work on which that information depends.”
BMC Medical Ethics (B)	Publisher (https://www.biomedcentral.com/getpublished/editorial-policies)	Permissive	“Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared.”
Cambridge Quarterly of Healthcare Ethics (B)	Publisher (https://www.cambridge.org/core/services/publishing-ethics/authorship-and-contributorship-journals#ai-contributions-to-research-content)	Permissive	“AI use must be declared and clearly explained in publications such as research papers, just as we expect scholars to do with other software, tools and methodologies.”
Canadian Journal of Bioethics (B)	No identifiable policy	–-	––
Christian Bioethics (B)	Journal (https://academic.oup.com/cb/pages/Instructions_To_Authors)	Non-permissive	“Additionally, authors must provide written confirmation that the paper is solely the work of the listed authors; neither ChatGPT nor any similar artificial intelligence program was used to assist in writing any part of the paper.”
Clinical Ethics (B)	Publisher (https://us.sagepub.com/en-us/nam/artificial-intelligence-policy)	Permissive	“We recognise that AI assisted writing has become more common as the technology becomes more accessible. AI tools that make suggestions to improve or enhance your own work, such as tools to improve language, grammar or structure, are considered assistive AI tools and do not require disclosure by authors or reviewers … The use of AI tools that can produce content such as generating references, text, images or any other form of content must be disclosed when used by authors or reviewers.”
Developing World Bioethics (B)	Publisher (https://authorservices.wiley.com/ethics-guidelines/index.html#22)	Permissive	“If an author has used a GenAI tool to develop any portion of a manuscript, its use must be described, transparently and in detail, in the Methods section (or via a disclosure or within the Acknowledgements section, as applicable). The author is fully responsible for the accuracy of any information provided by the tool and for correctly referencing any supporting work on which that information depends.”
Ethics, Medicine and Public Health (B)	Journal (https://www.sciencedirect.com/journal/ethics-medicine-and-public-health/publish/guide-for-authors)	Permissive	“Generative AI and AI-assisted technologies should only be used in the writing process to improve the readability and language of the manuscript.”
Global Bioethics (B)	Publisher (https://taylorandfrancis.com/our-policies/ai-policy/)	Permissive	“Authors must clearly acknowledge within the article or book any use of Generative AI tools through a statement which includes: the full name of the tool used (with version number), how it was used, and the reason for use. For article submissions, this statement must be included in the Methods or Acknowledgments section.”
Hastings Center Report (B)	Journal (https://onlinelibrary.wiley.com/page/journal/1552146x/homepage/forauthors.html)	Permissive	“Authors who have used large language models or other generative AI tools to develop or compose their paper should describe, in the cover letter and in a disclosure section or other relevant part of the manuscript, how in particular they used such tools. In describing the use of AI, err on the side of too much, rather than too little, transparency. AI tools cannot be listed as authors.”
Healthcare Analysis (B)	Publisher (https://www.springer.com/gp/editorial-policies/artificial-intelligence--ai-/25428500)	Permissive	“Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared.”
HEC Forum (B)	Publisher (https://www.springer.com/gp/editorial-policies/artificial-intelligence--ai-/25428500)	Permissive	“Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared.”
Hektoen International A Journal of Medical Humanities (HH)	No identifiable policy	–-	––
International Journal of Feminist Approaches to Bioethics (B)	No identifiable policy	–-	––
Journal of Bioethical Inquiry (B)	Publisher (https://www.springer.com/gp/editorial-policies/artificial-intelligence--ai-/25428500)	Permissive	“Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared.”
Journal of Empirical Research on Human Research Ethics (B)	Publisher (https://us.sagepub.com/en-us/nam/artificial-intelligence-policy)	Permissive	“We recognise that AI assisted writing has become more common as the technology becomes more accessible. AI tools that make suggestions to improve or enhance your own work, such as tools to improve language, grammar or structure, are considered assistive AI tools and do not require disclosure by authors or reviewers … The use of AI tools that can produce content such as generating references, text, images or any other form of content must be disclosed when used by authors or reviewers.”
Journal of Humanities in Rehabilitation (HH)	No identifiable policy	–-	––
Journal of Law, Medicine, and Ethics (B)	Publisher (https://www.cambridge.org/core/services/publishing-ethics/authorship-and-contributorship-journals#ai-contributions-to-research-content)	Permissive	“AI use must be declared and clearly explained in publications such as research papers, just as we expect scholars to do with other software, tools and methodologies.”
Journal of Literary & Cultural Disability Studies (HH)	Publisher (https://www.liverpooluniversitypress.co.uk/pb-assets/files/lup_publication_ethics-1686316566.pdf)	Permissive	“With regard to authorship and Artificial Intelligence (AI), Liverpool University Press agrees with COPE as outlined in their Position Statement.”
Journal of Medical Ethics (B)	Publisher (https://authors.bmj.com/policies/ai-use/)	Permissive	“BMJ will consider content where AI technologies are used. Our approach is one of transparency. Where AI technology has been used this should be clearly described.”
Journal of Medical Humanities (HH)	Publisher (https://www.springer.com/gp/editorial-policies/artificial-intelligence--ai-/25428500)	Permissive	“Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared.”
Journal of Medicine and Philosophy (B)	Journal (https://academic.oup.com/jmp/pages/Instructions_To_Authors)	Non-permissive	“Additionally, authors must provide written confirmation that the paper is solely the work of the listed authors; neither ChatGPT nor any similar artificial intelligence program was used to assist in writing any part of the paper.”
Journal of Pediatric Ethics (B)	No identifiable policy	–-	––
Kennedy Institute of Ethics Journal (B)	Journal (https://kiej.georgetown.edu/submit-to-the-kiej/)	Non-permissive	“*Authors submitting manuscripts to the KIEJ must confirm that they have not used generative AI to contribute to the content, style, or citations of their manuscripts. The Kennedy Institute of Ethics Journal reserves the right to later publicly retract published articles which are found to have breached this condition.”
Literature and Medicine (HH)	Publisher (https://www.press.jhu.edu/sites/default/files/media/2023/07/Generative%20AI%20for%20Authors_final_0.pdf)	Permissive	“Authors must disclose their use of AI or LLMs in the work they are submitting for possible publication.”
Medical Humanities (HH)	Publisher (https://authors.bmj.com/policies/ai-use/)	Permissive	“BMJ will consider content where AI technologies are used. Our approach is one of transparency. Where AI technology has been used this should be clearly described.”
Medicine, Health Care and Philosophy (B)	Publisher (https://www.springer.com/gp/editorial-policies/artificial-intelligence--ai-/25428500)	Permissive	“Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared.”
Neuroethics (B)	Publisher (https://www.springer.com/gp/editorial-policies/artificial-intelligence--ai-/25428500)	Permissive	“Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared.”
Nursing Ethics (B)	Publisher (https://us.sagepub.com/en-us/nam/artificial-intelligence-policy)	Permissive	“We recognise that AI assisted writing has become more common as the technology becomes more accessible. AI tools that make suggestions to improve or enhance your own work, such as tools to improve language, grammar or structure, are considered assistive AI tools and do not require disclosure by authors or reviewers … The use of AI tools that can produce content such as generating references, text, images or any other form of content must be disclosed when used by authors or reviewers.”
Pediatric Ethicscope (B)	No identifiable policy	–-	––
Pegasus Review: A Medical Literary Review Journal (HH)	No identifiable policy	–-	––
Perspectives in Biology and Medicine (B)	Publisher (https://www.press.jhu.edu/sites/default/files/media/2023/07/Generative%20AI%20for%20Authors_final_0.pdf)	Permissive	“Authors must disclose their use of AI or LLMs in the work they are submitting for possible publication.”
Public Health Ethics (B)	Publisher (https://academic.oup.com/pages/for-authors/books/author-use-of-artificial-intelligence)	Permissive	“AI does not qualify as an author and should not be used to undertake primary authorial responsibilities, such as generating arguments and scientific insights, writing analysis, or drawing conclusions. … Authors are responsible for the accuracy, integrity, and originality of their works, as well as any AI generated content these may include.”
Pulse-Voices from the Heart of Medicine (HH)	No identifiable policy	–-	––
Research & Humanities in Medical Education (RHiME) (HH)	Journal (https://www.rhime.in/ojs/index.php/rhime/about/submissions)	Permissive	“If they must use AI and AI-assisted technologies in the writing process, authors should use it only to improve readability of the text, and to fix syntax and grammar issues They must avoid using AI to generate hypothesis, draw conclusions, or make recommendations, since AI could potentially produce biased and incorrect content. Authors must take the final call (use human judgment) to decide whether to accept the suggestions made by AI or not, because, ultimately, the human author is accountable for the sanctity of the content they submit to RHiME Finally, in the interests of transparency and integrity, authors should make a disclosure in the beginning of the blinded article itself that they have used AI and AI-assisted technologies in their work. The disclosure will be published alongside the final work, if it is accepted for publication.”
Rhetoric of Health & Medicine (HH)	No identifiable policy	–-	––
Synapsis: A Health Humanities Journal (HH)	No identifiable policy	–-	––
The Healing Muse (HH)	No identifiable policy	–-	–-
The Postgraduate Journal of Medical Humanities (HH)	No identifiable policy	–-	––
The Journal of Clinical Ethics (B)	No identifiable policy	–-	––
The Journal of Hospital Ethics (B)	Journal (https://www.medstarhealth.org/locations/medstar-washington-hospital-center/the-john-j-lynch-md-center-for-ethics/the-journal-of-hospital-ethics)	Non-Permissive	“Any and all text generated by Artificial Intelligence (AI) will not be accepted for publication in JoHE. In addition, any accompanying AI-generated figures, images, and graphics cannot be used unless with explicit permission from JoHE’s Editorial Group.”
Theoretical Medicine & Bioethics (B)	Publisher (https://www.springer.com/gp/editorial-policies/artificial-intelligence--ai-/25428500)	Permissive	“Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared.”
Narrative Inquiry in Bioethics (B)	Publisher (https://www.press.jhu.edu/sites/default/files/media/2023/07/Generative%20AI%20for%20Authors_final_0.pdf)	Permissive	“Authors must disclose their use of AI or LLMs in the work they are submitting for possible publication.”
Voices in Bioethics (B)	Journal (https://journals.library.columbia.edu/index.php/bioethics/about/submissions)	Non-Permissive	“Voices in Bioethics does not accept AI-generated text including text generated by ChatGPT or any other LLM… Text attributed to such is not acceptable. Use of ChatGPT for text credited to the author is plagiarism.”​

A total of 10/15 editors responded to our follow-up email regarding their journal’s apparent lack of an AI policy. Each confirmed this lack of a policy, with 7 stating that they are in active discussion with other members of the journal editorial team about developing a policy and 3 stating that they would not want to receive an AI-assisted manuscript.

Discussion

Our study indicates a lack of clear guidance for bioethicists and health/medical humanists on journal homepages (e.g., author guidelines, policy statements, submission requirements), which confirms research across academic disciplines [9, 22, 23]. For authors who may use AI to assist in the writing process, the lack of clear AI guidance can be confusing and frustrating. While a publisher may have a statement on AI use, potential authors may be unclear about whether said policy applies to one of its journals, especially if said journal does not clearly affirm the policy. If authors disclose AI utilization in manuscript preparation, they might worry that journal editors will reject their work based on unwritten or inconsistently applied rules (see [7]). Conversely, if authors choose not to disclose AI use due to uncertainty, they may later face ethical concerns or retraction risks if AI involvement is discovered and deemed inappropriate. Furthermore, the absence of transparent AI policies can contribute to inconsistencies in editorial decision-making, as different editors within the same journal may apply varying interpretations of AI acceptability. The lack of guidance also raises broader ethical and scholarly integrity concerns. AI-assisted writing spans a spectrum, from grammar and style suggestions to more substantial content generation, yet journals without explicit policies leave authors guessing about what level of AI involvement is acceptable.

Our study confirms broad consensus around the following practices [13, 24, 34]. First, AI does not meet the standards of authorship and accordingly cannot be listed as an author. Authorship in academic publishing, per ICMJE [15] guidelines, is based on intellectual contribution, accountability, and the ability to take responsibility for a manuscript’s content. Since AI lacks independent reasoning, intent, and accountability, it does not fulfill these criteria. Second, authors are responsible for the content of their submission. Even if AI tools are used in manuscript preparation, human authors bear full accountability for the accuracy, originality, and ethical integrity of their work. Editors and peer reviewers evaluate manuscripts under the expectation that authorship implies ownership and oversight, which cannot be delegated to AI. Third, authors should declare their use of AI to the editors as well as acknowledge it in the manuscript itself. Transparency in AI usage is essential to maintaining integrity in scholarly communication. Fourth, reliance upon AI should not be substantial, meaning that authors should not rely on it to generate ideas or write the manuscript. While AI can assist with tasks such as grammar refinement, language translation, or literature summarization, it should not replace the critical thinking, analytical insight, and originality expected of scholarly authorship.

Among AI policies we examined, there is noteworthy disagreement among journals and publishers that have favorable AI policies [20, 32]. First, there is disagreement around the extent of AI-assistance that can be considered. The publisher Sage [35] requires authors to disclose to the Editorial team if a manuscript “was primarily or partially generated using AI…so that the Editorial team can evaluate the content generated.” This suggests an openness of Sage journals to considering manuscripts that were substantially generated by AI. Taylor and Francis [37] cautions that “some [of their] journals may not allow use of Generative AI tools beyond language improvement”, which suggests ambiguity among their journals regarding the extent of AI-assistance is accepted. John Hopkins University Press [17] affirms that it “will not accept any work that is substantially written by an AI or LLM tool.” Second, there is disagreement whether every use of AI requires disclosure. The publisher Springer Nature [36] posits that AI-assisted “copy editing” (i.e., using AI to improve style, grammar, and readability) “does not need to be declared.” Cambridge University Press [5], by contrast, posits that “AI use must be declared and clearly explained in publications such as research papers, just as we expect scholars to do with other software, tools and methodologies.” John Hopkins University Press [17] state that: “Authors must disclose their use of AI or LLMs in the work they are submitting for possible publication.” Therefore, researchers should ensure that they familiarize themselves with journal policy prior to submitting a manuscript that was supported by AI.

Some of the journals that do not allow AI utilization in any capacity do not explain their rationale. At both Christian Bioethics and the Journal of Philosophy and Medicine, prospective authors are required to “provide written confirmation that the paper is solely the work of the listed authors; neither ChatGPT nor any similar artificial intelligence program was used to assist in writing any part of the paper.” The Kennedy Institute of Ethics Journal similarly requires authors to “confirm that they have not used generative AI to contribute to the content, style, or citations of their manuscripts.” The journal Voices in Bioethics prohibits AI-generated text on the grounds that it constitutes plagiarism. However, whether AI-assisted writing qualifies as plagiarism is debatable, as plagiarism traditionally involves presenting another person’s ideas or words as one’s own. Since AI does not possess personhood, the prevailing view is that its utilization does not amount to plagiarism [16]. It is important for journals that refuse to consider AI-assisted manuscripts to articulate a clear rationale for doing so. Potential reasons include the importance of generating original scholarship, the concern that AI-use fosters deskilling, and the dangers of epistemic risk [10, 25].

The feedback from journals and editors of journals without an AI policy confirms the broad, growing awareness among editors and journals of the value of such a policy [20, 32]. As scholars increasingly utilize AI in the research and writing process [26, 27], journals are increasingly likely to receive submissions that were assisted to some extent by AI: for better or worse, AI-assistance is here to stay, and so it is important for journals to develop clear guidelines. There are challenges to acting on an AI policy, given how difficult it can be to determine AI–generated content [14]. Still, it is better to have a clear policy in place than not, and the feedback we received confirms that editors are aware of this need.

There is a tension that is worth noting. An editor of two journals that explicitly disallows AI-assistance signed the joint statement on the use of AI in scholarly writing [19]. In addition, both of the journals are published by Oxford University Press [31], which requires that AI use “be disclosed both in cover letters to editors and in the Methods or Acknowledgements section of manuscripts”—this suggests that the publisher is receptive to AI assisted manuscripts. While this tension is curious, it reinforces the need for authors to carefully read the journal submission guidelines to discern potential AI policy. What matters most when submitting to a particular journal are the journal’s specific author instructions.

Limitations

This study does have limitations. The study’s findings are limited to the specific journals included in the analysis. The selection process, while utilizing reputable sources, may not have captured all relevant journals in the field of bioethics and health/medical humanities. In addition, the number of journals without explicit AI policies makes it difficult to fully assess the field’s stance on AI. The absence of publicly available policies does not account for informal practices, such as internal editorial discussions or unpublished operational guidelines.

Furthermore, the extent to which journal editors adhere to publisher-level AI policies remains unexamined. The assumption that a journal without an explicit AI policy defaults to its publisher’s policy is contestable, as editors may not actively enforce or endorse these guidelines. Factors such as editorial bias or implicit resistance may influence policy implementation. Future research should explore editors’ perspectives on AI-assisted manuscripts to provide a more comprehensive understanding of AI governance in scholarly publishing.

Conclusion

AI policies are rapidly evolving, and this study presents one snapshot in time. As AI continues to evolve and become more embedded in academic writing, the need for transparent, standardized policies becomes increasingly critical to ensure fairness, clarity, and consistency in scholarly publishing. One especially important challenge is determining the appropriate threshold at which AI utilization warrants disclosure (see [33]). The time is therefore ripe for a professional society such as the American Society of Bioethics and Humanities to consider forming an expert group tasked with developing consensus guidelines on the use of AI within bioethics and health humanities publishing.

Abbreviations

AI

Artificial intelligence

LLM

Large language model

ICMJE

International Committee of Medical Journal Editors

Authors’ contributions

CB designed the study and drafted the manuscript; DR and DH reviewed findings and contributed to the writing of the manuscript. ChatGPT was used to improve word processing and language of a few sentences in the discussion section in addition to generating the figures presenting the research findings. All three authors affirm the manuscript’s content. CB is the guarantor.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Data availability

The data are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.