CoChat
Sign Up
Cited by 266Academic integrity and plagiarismEthics and Social Impacts of AIAcademic Writing and Publishing

The ethics of disclosing the use of artificial intelligence tools in writing scholarly manuscripts

Open access
Hosseini, Mohammad, Resnik, David B, Holmes, Kristi·2023·Research Ethics
DOI 10.1177/17470161231180449PMC PMC11694804
Open this paper in CoChat — free accountOpen PDF

Abstract

In this article, we discuss ethical issues related to using and disclosing artificial intelligence (AI) tools, such as ChatGPT and other systems based on large language models (LLMs), to write or edit scholarly manuscripts. Some journals, such as Science , have banned the use of LLMs because of the ethical problems they raise concerning responsible authorship. We argue that this is not a reasonable response to the moral conundrums created by the use of LLMs because bans are unenforceable and would encourage undisclosed use of LLMs. Furthermore, LLMs can be useful in writing, reviewing and editing text, and promote equity in science. Others have argued that LLMs should be mentioned in the acknowledgments since they do not meet all the authorship criteria. We argue that naming LLMs as authors or mentioning them in the acknowledgments are both inappropriate forms of recognition because LLMs do not have free will and therefore cannot be held morally or legally responsible for what they do. Tools in general, and software in particular, are usually cited in-text, followed by being mentioned in the references. We provide suggestions to improve APA Style for referencing ChatGPT to specifically indicate the contributor who used LLMs (because interactions are stored on personal user accounts), the used version and model (because the same version could use different language models and generate dissimilar responses, e.g., ChatGPT May 12 Version GPT3.5 or GPT4), and the time of usage (because LLMs evolve fast and generate dissimilar responses over time). We recommend that researchers who use LLMs: (1) disclose their use in the introduction or methods section to transparently describe details such as used prompts and note which parts of the text are affected, (2) use in-text citations and references (to recognize their used applications and improve findability and indexing), and (3) record and submit their relevant interactions with LLMs as supplementary material or appendices.

AI snapshot

Written from the paper’s full text.

TL;DR

The paper argues that banning LLMs in scholarly manuscript writing is unreasonable because bans are unenforceable and would encourage undisclosed use, while LLMs can support equity in science. It contends LLMs should not be authors or acknowledged, but should be disclosed via free text, in-text citations/references, and supplementary materials, with a modified APA citation format.

Key findings

  • The authors oppose journal bans on LLMs for three reasons: bans are unenforceable, they may encourage undisclosed use, and LLMs can help researchers who are not highly proficient in English and thus promote equity in science.
  • LLMs should not be named as authors or mentioned in acknowledgments because they lack free will and cannot be held morally or legally responsible, cannot consent, and cannot approve a final manuscript. The paper discusses examples: an O’Connor editorial in *Nurse Education in Practice* listed ChatGPT as second author, with ChatGPT drafting five out of seven paragraphs; a Zhavoronkov *Oncoscience* paper listed ChatGPT as co-author and sought approval from OpenAI CEO Sam Altman, which the authors reject as insufficient.
  • The authors recommend disclosing LLM use through (1) free text in the introduction or methods section describing who used LLMs, when, how, prompts, and affected text; (2) in-text citations and references; and (3) recorded interactions as supplementary material or appendices. They suggest a modified APA reference format: “OpenAI (2023). ChatGPT (GPT-4, May 12 Version) [Large language model]. Response to query made by X.Y. Month/Day/Year. https://chat.openai.com/chat”.
  • The paper cites Blanco-Gonzalez et al. (2022), who reported that 33.9% of their manuscript comprised ChatGPT-generated text: identical 4.3%, minor changes 13.3%, and related meaning 16.3%. It also notes that ChatGPT May 12 Version can use different models (GPT-3.5 or GPT-4) and generate dissimilar responses.

Methods

  • Conceptual/normative ethics analysis; no empirical sample, systematic review, or statistical analysis is reported.
  • The argument draws on published journal editorials and policies (e.g., *Science*, *Nature*, *JAMA*, *Accountability in Research*, WAME, COPE, ICMJE, APA) and illustrative cases of LLM authorship/use.
  • Main analysis is philosophical and policy-oriented: it examines free will, responsibility, accountability, consent, and credit allocation, then proposes disclosure best practices and a modified citation format.

Limitations

  • Stated in the paper: ensuring consistency of disclosures could be challenging; disclosure only in text raises findability/indexing and paywall-access problems; detailed quantification of LLM contribution is unlikely to be provided consistently and may be impossible for non-quantifiable tasks such as conceptualization.
  • Stated in the paper: bans are unenforceable partly because detecting LLM-generated text is extremely difficult and detection tools are unreliable.
  • Evident from the text: the recommendations are normative and not empirically tested; the examples are illustrative rather than systematic; the paper does not engage the metaphysical free will debate and offers no data on prevalence or effectiveness of disclosure practices.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1177/17470161231180449

Cites

32 of 32 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

10 of 54 cited sources checked · 1 supported · 0 flagged · accuracy not yet measured

  • Premisecitation

    LLMs can be useful in writing, reviewing and editing text.

    “Furthermore, LLMs can be useful in writing, reviewing and editing text, and promote equity in science.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Hosseini and Horbach, 2023 source
  • Interpretationcitation

    Not disclosing LLMs, especially those which provide context-specific suggestions and can generate or substantially affect content, violates norms of ethical attribution because it results in giving undue credit to human contributors for work which they did not do.

    “With respect to fair allocation of credit, not disclosing LLMs, especially those which provide context-specific suggestions and can generate or substantially affect content, violates norms of ethical attribution because it results in giving undue credit to (human) contributors for work which they did not do (Verhoeven et al., 2023).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Verhoeven et al., 2023 source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
  • Premisecitation

    Some journals, such as Science, have banned the use of LLMs because of the ethical problems they raise concerning responsible authorship.

    “Some journals, such as Science, have banned the use of LLMs because of the ethical problems they raise concerning responsible authorship.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Thorp, 2023 source

      The policy passage says Science journals prohibit ChatGPT- or other AI-generated text and graphics in papers, and that an AI program cannot be an author. It links those rules to originality and authors’ accountability. But it does not ban all use of LLMs: it expressly exempts AI-generated research datasets. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
  • Premisecitation

    Accountability in Research, JAMA, and Nature decided to adopt or pursue policies that allow using LLMs under conditions that promote transparency, accountability, fair assignment of credit, and honesty.

    “While Accountability in Research, JAMA, and Nature decided to adopt or pursue policies that allow using LLMs under conditions that promote transparency, accountability, fair assignment of credit, and honesty, the editors of Science highlighted ethical problems created by LLMs and banned their use:”
    Quote in the paperNot a substantial claim2 of 3 sources checked
    • cites Hosseini et al., 2023 source

      The Social and epistemic impacts of peer review section says several journals had attempted to develop LLM guidelines, and the Recommendations call for disclosure and human responsibility. Our copy does not say that Accountability in Research, JAMA, and Nature each adopted or pursued policies allowing LLM use under the stated conditions. The retrieved passages have section labels but no page numbers. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites Flanagin et al., 2023 source

      In the main text (page unavailable), the paper says JAMA updated its policies to require disclosure of AI-assisted writing and describes Nature’s policy requiring disclosure while barring AI tools from authorship credit. It does not identify Accountability in Research or establish its policy, and it does not explicitly describe honesty as a condition of these policies. (LLM, conf=0.85)

      Experimental — accuracy not yet measured
    • cites Nature, 2023

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Premisecitation

    LLMs can play an important role in helping researchers who are not highly proficient in English to write and edit their papers, or review others' manuscripts.

    “Third, LLMs can play an important role in helping researchers who are not highly proficient in English (the lingua franca for most top journals) to write and edit their papers, or review others’ manuscripts (Hosseini and Horbach, 2023), which could promote equity in science (Berdejo-Espinola and Amano, 2023).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Hosseini and Horbach, 2023 source

      The Roles of editors in peer review section says LLMs could help non-native English speakers turn their views into presentable reviews. The Roles of reviewers in peer review section also describes manuscript editing and mentions using LLMs to write manuscripts, but does not specifically connect those authoring benefits to researchers with limited English proficiency. Page numbers are unavailable in our copy. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
  • Classified as probably not a substantial claim. 42
  • May combine two claims. 19

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    LLMs can make some glaring mistakes, are susceptible to bias, and may even fabricate facts or citations.

    “Although LLMs can make some glaring mistakes, are susceptible to bias, and may even fabricate facts or citations (Hosseini et al., 2023), these flaws should not be held against them because human researchers might make similar errors.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Hosseini et al., 2023 source

      The Recommendations section warns that LLM-assisted reviews may be biased or outright erroneous, and the Roles of reviewers section describes a biased ChatGPT response. Our copy does not establish the more specific claim that LLMs fabricate facts or citations; searches for fabricated or invented references did not find such a passage. The citing authors’ argument that similar human errors excuse these flaws is not attributed to this source. (LLM, conf=0.85)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Premisecitation

    Accountability and credit are two sides of the same coin, and contributors cannot have one without the other.

    “Accountability and credit are two sides of the same coin, and contributors cannot have one without the other (Hosseini et al., 2022; Resnik, 1997; Smith, 2017).”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites Hosseini et al., 2022 source

      The “Attribution of Credit” and “Responsibility and Accountability” sections link recognition and accountability to identifiable contributions, and say authors should be accountable for their work. They do not establish the stronger claim that every contributor must have credit and accountability together; the paper distinguishes authorship credit from acknowledgement for different contributions. Page numbers are unavailable in our copy. (LLM, conf=0.83)

      Experimental — accuracy not yet measured
    • cites Resnik, 1997 source
    • cites Smith, 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Tools used in research are typically disclosed in the body of the text, and software applications are also cited with in-text citations and in the references.

    “Tools used in research are typically disclosed in the body of the text, and in the case of software applications, they are also cited with in-text citations and in the references (Katz et al., 2021).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Katz et al., 2021 source

      The Software citation essentials section recommends citing software used in research, including software described in the Methods section. The In-text referencing and Software citation examples sections show in-text citations and reference entries for software. The paper offers guidance and examples, however; it does not establish that research tools are typically disclosed or that software is routinely cited this way. (LLM, conf=0.84)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Mentioning LLMs in the acknowledgments is an inappropriate form of recognition.

    “We argue that naming LLMs as authors or mentioning them in the acknowledgments are both inappropriate forms of recognition because LLMs do not have free will and therefore cannot be held morally or legally responsible for what they do.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Premisecitation

    LLMs can promote equity in science.

    “Furthermore, LLMs can be useful in writing, reviewing and editing text, and promote equity in science.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Berdejo-Espinola and Amano, 2023 source

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Naming LLMs as authors is an inappropriate form of recognition.

    “We argue that naming LLMs as authors or mentioning them in the acknowledgments are both inappropriate forms of recognition because LLMs do not have free will and therefore cannot be held morally or legally responsible for what they do.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Premisecitation

    One can be held morally and legally responsible for an action only if that action results from one's free choices.

    “One can be held morally and legally responsible for an action only if that action results from one’s free choices (Mele, 2006).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mele, 2006 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The O'Connor editorial lists ChatGPT as the second author.

    “For example, in an editorial titled “Open artificial intelligence platforms in nursing education: Tools for academic progress or abuse?” published in the journal of Nurse Education in Practice, ChatGPT is listed as the second author (O’Connor, 2023).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites O’Connor, 2023

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Both challenges (findability of articles that used LLMs and identifying what part of the text is affected by their use) could be resolved via general norms of software citation that include in-text citations and referencing.

    “Both challenges (i.e. findability of articles that used LLMs and identifying what part of the text is affected by their use) could be resolved via general norms of software citation that include in-text citations and referencing.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationreasoning

    Because LLMs cannot be held morally or legally responsible, they also cannot be held accountable for their actions.

    “Because LLMs cannot be held morally or legally responsible, they also cannot be held accountable for their actions.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Approval from the CEO of OpenAI should not be considered approval from the author.

    “However, approval from the CEO of OpenAI should not be considered approval from the author, any more than approval by the corresponding author of a paper should count as approval by other (human) authors.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

    Premisenone given

    LLMs learn fast and change rapidly.

    “Let’s not forget that LLMs learn fast and change rapidly, making it vital to disclose not only what version was used but also which model, when and by who.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

    Interpretationcitation

    Ensuring the consistency of disclosures could be challenging and could be addressed through training and education, as well as promoting best practices.

    “That said, ensuring the consistency of disclosures could be challenging (similar challenges are faced in software citation, e.g. see Li et al., 2017) and could be addressed through training and education, as well as promoting best practices.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Li et al., 2017

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationreasoning

    Being able to explain what they have done and why would take LLMs a step closer to being accountable but would still fall far short of the degree of accountability expected from human beings.

    “Although being able to explain what they have done and why would take LLMs a step closer to being accountable, it would still fall far short of the degree of accountability we expect from human beings.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisereasoning

    Posthumous authorship is generally granted for works that individuals contributed to and would endorse if they were alive.

    “While it is possible that this sometimes might not be the case (e.g. a work might be irresponsibly or even maliciously attributed to someone who would have not agreed to be an author if they were alive), generally posthumous authorship is granted for works that individuals contributed to, and would endorse if they were alive.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisereasoning

    Detecting text that has been generated with LLMs is extremely difficult, partly because LLM-generated text can be altered by human beings to mask it.

    “Even if prominent research institutions and journals were to adopt such measures, these efforts would likely be in vain, since detecting text that has been generated with LLMs is extremely difficult, partly because LLM-generated text can be altered by human beings to mask it.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

    Interpretationreasoning

    Use of LLMs should be cited in-text, followed by being mentioned in the references.

    “Use of LLMs, like the use of other type of software tools, should be cited in-text, followed by being mentioned in the references.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Tools in general, and software in particular, are usually cited in-text, followed by being mentioned in the references.

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

    Interpretationreasoning

    Since LLMs are not persons, the social aspects of authorship do not apply to them.

    “Since LLMs are not persons, these social aspects of authorship do not apply to them, but they do apply to human authors, even dead ones.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Premisecitation

    Authorship is a form of social capital that is awarded, protected, and exchanged in human relationships.

    “Authorship, in this sense, is a form of social capital that is awarded, protected, and exchanged in human relationships (Smith, 2017).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Smith, 2017 source

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationreasoning

    LLMs cannot consent to being acknowledged.

    “We believe that crediting LLMs in the acknowledgments section of a manuscript is inappropriate for largely the same reasons that LLMs should not be named as authors, that is, because they lack free will and therefore cannot consent to being acknowledged.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationnone given

    LLMs lack free will.

    “We believe that crediting LLMs in the acknowledgments section of a manuscript is inappropriate for largely the same reasons that LLMs should not be named as authors, that is, because they lack free will and therefore cannot consent to being acknowledged.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationcitation

    Ethical principles, including openness, honesty, transparency, efficient use of resources, and fair allocation of credit, demand disclosing the use of LLMs.

    “Ethical principles, including openness, honesty, transparency, efficient use of resources, and fair allocation of credit (Shamoo and Resnik, 2022) demand disclosing the use of LLMs.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Shamoo and Resnik, 2022 source
    • Rests on an assumption: Openness, transparency, and honesty about used methods and tools are paramount to fostering integrity, reproducibility, and rigor in research.

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationreasoning

    These flaws should not be held against LLMs because human researchers might make similar errors.

    “Although LLMs can make some glaring mistakes, are susceptible to bias, and may even fabricate facts or citations (Hosseini et al., 2023), these flaws should not be held against them because human researchers might make similar errors.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

    Interpretationnone given

    If AIs develop to the point where there is compelling evidence that they have free will and can be held responsible and accountable and can participate in society like humans, then they could be named as authors on scholarly publications.

    “If AIs develop to the point where there is compelling evidence that they have free will and can be held responsible and accountable and can participate in society like humans, then they could be named as authors on scholarly publications.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Corporations, government entities, and research centers can own copyrights, can be held morally and legally responsible and accountable, and can approve the final version of a manuscript.

    “However, this objection is misleading because corporations, government entities, and research centers can own copyrights,3 can be held morally and legally responsible and accountable and can even approve the final version of a manuscript.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites ICMJE, 2023

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Posthumous authors were capable of being held accountable and of approving the final version of the paper when they did the work they are credited for.

    “However, we believe the ethical acceptability of posthumous authorship is not a convincing objection to our view because posthumous authors were capable of being held accountable and of approving the final version of the paper when they did the work they are credited for.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

    Interpretationreasoning

    Today's LLMs are neither responsible nor accountable because they lack free will.

    “Today’s LLMs are neither responsible nor accountable because they lack free will (or self-determination).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Current LLMs lack consciousness, self-awareness, a humanlike understanding of language, and values or preferences.

    “While they can manipulate linguistic symbols and digital data quite adeptly, they lack consciousness, self-awareness, a humanlike understanding of language, and values or preferences (Bogost, 2022; Teng, 2020).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Bogost, 2022

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Teng, 2020

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Premisecitation

    For an individual to make a deliberate choice, they must have consciousness, self-awareness, understanding, the ability to reason, information, and values or preferences.

    “For an individual to make a deliberate choice, they must have consciousness, self-awareness, understanding, the ability to reason, information, and values or preferences (see Mele, 2006; O’Connor, 2022).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Mele, 2006 source
    • cites O’Connor, 2022

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim.

    Premisecitation

    APA advises disclosure in the methods section in research articles or in the introduction in literature reviews, essays or response or reaction papers.

    “APA advises disclosure in the methods section in research articles or in the introduction in literature reviews, essays or response or reaction papers (McAdoo, 2023).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites McAdoo, 2023

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationreasoning

    LLM-detection tools are unreliable and are likely to remain unreliable in finding LLM-generated text.

    “Although some companies, including OpenAI, have developed software designed to recognize LLM-generated text (Hu, 2023), these tools are unreliable and are likely to remain unreliable in finding LLM-generated text as computer scientists and researchers find ways of working around them.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Hu, 2023

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    In manuscripts behind a paywall, citations are not accessible to all readers, but corresponding references are often open, and thanks to open citations initiatives will likely become more accessible.

    “In manuscripts behind a paywall, citations are not accessible to all readers, but corresponding references are often open, and thanks to open citations initiatives (e.g. I4OC) will likely become more accessible.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites I4OC

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    In-text citations offer the required signposting to indicate what part of the text is affected by LLMs.

    “Indeed, in-text citations offer the required signposting to indicate what part of the text is affected by LLMs.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Premisecitation

    In research, accountability is essential for promoting integrity, reproducibility, rigor, and other important epistemic and moral values.

    “In research, accountability is essential for promoting integrity, reproducibility, rigor, and other important epistemic and moral values (Shamoo and Resnik, 2022).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Shamoo and Resnik, 2022 source

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationnone given

    LLMs cannot be considered responsible or accountable in the way human beings can be.

    “However, as mentioned earlier, recognition is not only about assigning credit but also involves responsibility and accountability, and LLMs cannot be considered responsible or accountable in the way human beings can be.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationreasoning

    The ethical acceptability of posthumous authorship is not a convincing objection to the view that LLMs should not be named as authors.

    “However, we believe the ethical acceptability of posthumous authorship is not a convincing objection to our view because posthumous authors were capable of being held accountable and of approving the final version of the paper when they did the work they are credited for.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Interpretationcitation

    LLMs should not be named as authors because they cannot be held legally and morally responsible for what they do, and authorship implies responsibility.

    “In summary, LLMs should not be named as authors because they cannot be held legally and morally responsible for what they do, and authorship implies responsibility (Copyright Review Board, 2022; Shamoo and Resnik, 2022).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Copyright Review Board, 2022

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Shamoo and Resnik, 2022 source

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

    Interpretationreasoning

    LLMs cannot be held morally or legally responsible for what they do.

    “because LLMs do not have free will and therefore cannot be held morally or legally responsible for what they do.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisenone given

    LLMs do not have interests, values, or feelings.

    “These and other forms of punishment play an important role in deterring misconduct in research (Horner and Minifie, 2011), but punishments cannot affect (let alone deter) LLMs in any way, because they do not have interests, values, or feelings.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Interpretationreasoning

    The level of detail about the extent of LLM use (e.g., percentage of text generated) is unlikely to be provided consistently by all researchers and is perhaps impossible to calculate when LLMs contribute to non-quantifiable tasks such as conceptualization.

    “This level of detail is unlikely to be provided consistently by all researchers and is perhaps impossible to calculate when LLMs contribute to tasks that are not quantifiable, such as conceptualization.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationreasoning

    When LLMs are used to edit and rewrite manuscripts, they are likely to generate or substantially affect content.

    “When LLMs are used to edit and rewrite manuscripts, they are likely to generate or substantially affect content.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisenone given

    According to the developers, different versions of ChatGPT (GPT-3.5 and GPT-4) offer different degrees of reasoning, speed, and conciseness, and provide dissimilar responses to the same prompt.

    “According to the developers, each of these versions offers different degrees of reasoning, speed, and conciseness, but more importantly, they provide dissimilar responses to the same prompt.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    LLM use in existing word processing programs such as MS Word is likely to happen soon.

    “One might also ask whether LLM use should be disclosed if they are incorporated into existing word processing programs, such as MS Word, which is likely to happen soon (Kelly, 2023).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Kelly, 2023

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Banning LLMs would not support diversity and inclusion in scholarship.

    “Also, since LLMs can have some useful applications in writing and editing text (especially for those conducting research in a language other than their first language), banning them would not support diversity and inclusion in scholarship.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    Banning the use of LLMs would encourage undisclosed use of LLMs.

    “We argue that banning the use of LLMs would be a mistake because a ban would not be enforceable and would encourage undisclosed use of LLMs.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationnone given

    LLMs can have some useful applications in writing and editing text, especially for those conducting research in a language other than their first language.

    “Also, since LLMs can have some useful applications in writing and editing text (especially for those conducting research in a language other than their first language), banning them would not support diversity and inclusion in scholarship.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationreasoning

    The situation where LLMs are only used in ways that do not generate or substantially affect content will be rare because LLMs can do so much more than correct grammatical or typographical errors.

    “While we think it is not necessary to disclose the use of LLMs if they are only used in ways that do not generate or substantially affect content, we think that this situation will be rare because LLMs can do so much more than correct grammatical or typographical errors.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: LLMs can do so much more than correct grammatical or typographical errors.

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationnone given

    Banning the use of LLMs would not be enforceable.

    “We argue that banning the use of LLMs would be a mistake because a ban would not be enforceable and would encourage undisclosed use of LLMs.”
    Quote in the paperNot a substantial claim
48 of 48

  • The paper relies on this without support.

    Openness, transparency, and honesty about used methods and tools are paramount to fostering integrity, reproducibility, and rigor in research.

    “Openness, transparency, and honesty about used methods and tools are paramount to fostering integrity, reproducibility, and rigor in research.”

  • The paper relies on this without support.

    Bans on LLM use may encourage undisclosed use of LLMs.

    “Second, bans may encourage undisclosed use of LLMs, which would undermine transparency and integrity in research and discourage training and education in responsible use of LLMs.”

  • The paper relies on this without support.

    Tools in general, and software in particular, are usually cited in-text, followed by being mentioned in the references.

    “Tools in general, and software in particular, are usually cited in-text, followed by being mentioned in the references.”

  • The paper relies on this without support.

    Approval is a form of consent and one cannot consent to something without free will.

    “However, it did not approve the final version of the manuscript because approval is a form of consent and one cannot consent to something without free will, which ChatGPT and other LLMs do not currently have, despite some sensational claims to the contrary (de Cosmo, 2022).”

  • The paper relies on this without support.

    Researchers do not mention search engines, scholarly indices, or software like SPSS in the acknowledgments section.

    “For example, researchers use search engines, such as Google, and scholarly indices, such as PubMed and Web of Science, to search the extant literature and find resources or use software like SPSS to identify relationships and correlations, but none of these tools are mentioned in the acknowledgments section.”

  • The paper relies on this without support.

    As of May 2023, ChatGPT PLUS allows choosing between GPT-3.5 and GPT-4 from the same version.

    “As per May 2023, when using ChatGPT PLUS (the paid version), one can choose between two different models (GPT-3.5 and GPT-4) from the same version (ChatGPT May 12 Version) to generate text.”

  • The paper relies on this without support.

    ChatGPT can store previous interactions on individual user accounts.

    “Especially in systems like ChatGPT that can generate dissimilar responses to similar prompts, and also store previous interactions on individual user accounts, collecting this information is required to ensure openness and transparency.”

  • The paper relies on this without support.

    A person may not want to be mentioned in the acknowledgments section if they disagree with the conclusions of a study.

    “Moreover, a person may not want to be mentioned in the acknowledgments section if they disagree with the conclusions of a study and do not want to be associated with it.”

  • The paper relies on this without support.

    A mention in the acknowledgments section carries moral and legal weight.

    “Although a mention in the acknowledgments section of a paper is not as prestigious as an author byline, it still carries some moral and legal weight and should therefore involve consent.”

  • The paper relies on this without support.

    People mentioned in the acknowledgments section may be held morally or legally responsible for the integrity of data they provided.

    “If a person mentioned in the acknowledgments section provided data for a study that comes under suspicion of fabrication or falsification, they may be held morally (and perhaps legally) responsible and accountable for the integrity of the data they provided.”

  • The paper relies on this without support.

    LLMs can do so much more than correct grammatical or typographical errors.

    “we think that this situation will be rare because LLMs can do so much more than correct grammatical or typographical errors.”
11 of 11

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?

  • The need for guidance on AI use in scientific research is urgent because using AI raises novel epistemological and ethical issues related to objectivity, reproducibility, transparency, accountability, responsibility, and trust in science.

    “Moreover, the need for guidance is urgent because using AI raises novel epistemological and ethical issues related to objectivity, reproducibility, transparency, accountability, responsibility, and trust in science [9, 102].”

    From The ethics of using artificial intelligence in scientific research: new guidance needed for a new tool (2025) · cites this paper as [102]

  • In the winter of 2022-2023, some researchers listed ChatGPT as authors on papers, making AI moral agency a major issue for ethical norms of science.

    “This became a major issue for the ethical norms of science in the winter of 2022–2023, when some researchers listed ChatGPT as authors on papers [102].”

    From The ethics of using artificial intelligence in scientific research: new guidance needed for a new tool (2025) · cites this paper as [102]

  • There is an emerging consensus that AI systems can be used in preparing manuscripts but that their use should be appropriately disclosed and discussed.

    “Currently, there seems to be an emerging consensus that falls in between these two extremes position, namely, that AI systems can be used in preparing manuscripts but that their use should be appropriately disclosed and discussed, [4, 102].”

    From The ethics of using artificial intelligence in scientific research: new guidance needed for a new tool (2025) · cites this paper as [102]

Cites

Flagged first, then by citations

Research works printed but not linked: 6

  1. Copyright Review Board (2022) Re: Second Request for Reconsideration for Refusal to Register A Recent Entrance to Paradise (Correspondence ID 1-3ZPC6C3
  2. ICML 2023 (n.d) International Conference on Machine Learning - ICML. Available at: https://icml.cc/Conferences/2023/CallForPapers (accessed 12 April 2023)
  3. International Committee of Medical Journal Editors (2023) Defining the Role of Authors and Contributors. https://www.icmje.org/recommendations/browse/manuscript-preparation/preparing-for-submission.html (accessed 10 April 2023)
  4. Mennell R, Burr S (2017) Wills and Trusts. St. Paul, MN: West Publishing.
  5. Miller A, Davis M (2018) Intellectual Property. St. Paul, MN: West Publishing.
  6. Nurse Education in Practice (2023) Guide for authors. Available at: https://www.elsevier.com/journals/nurse-education-in-practice/1471-5953/guide-for-authors (accessed 1 April 2023)

Cited by

Sorted by citations

Related

Authors
Hosseini, M., Resnik, D. B., & Holmes, K.
Affiliations
  • Northwestern University (US)
  • National Institute of Environmental Health Sciences (US)
Corresponding author
Hosseini, Mohammad
ORCID
Venue
Research Ethics
Publisher
SAGE Publications
Published
2023-06-15
Volume
19
Issue
4
Pages
449-465
ISSN
  • 1747-0161 (ISSN-L)
  • 2047-6094
DOI
10.1177/17470161231180449
PMC
PMC11694804
OpenAlex
W4380997384
Cited by
253
References
28
Citations by year
2026: 93 · 2025: 81 · 2024: 61 · 2023: 20
FWCI
64.53
Type
journal article
Language
en
Open access
Licence
Locations
  • Research Ethics (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • PubMed Central (submittedVersion, cc-by-nc)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Center for Advancing Translational Sciences (UL1TR001422)
  • National Institute of Environmental Health Sciences
Topics
  • Social Sciences › Safety Research › Academic integrity and plagiarism
  • Social Sciences › Safety Research › Ethics and Social Impacts of AI
  • Social Sciences › Arts and Humanities › History and Philosophy of Science › Academic Writing and Publishing
Keywords
large language models, authorship ethics, AI disclosure, author responsibility, research transparency, equity in science

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.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-29.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-29.

Cited sources are checked by gpt-6-sol.

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