CoChat
Sign Up
Cited by 140Academic integrity and plagiarismArtificial Intelligence in Healthcare and EducationAcademic Writing and Publishing

Artificial intelligence-assisted academic writing: recommendations for ethical use

Open access
Cheng, Adam, Calhoun, Aaron, Reedy, Gabriel·2025·Advances in Simulation
DOI 10.1186/s41077-025-00350-6PMC PMC12007126
Open this paper in CoChat — free accountOpen PDF

Abstract

Generative artificial intelligence (AI) tools have been selectively adopted across the academic community to help researchers complete tasks in a more efficient manner. The widespread release of the Chat Generative Pre-trained Transformer (ChatGPT) platform in 2022 has made these tools more accessible to scholars around the world. Despite their tremendous potential, studies have uncovered that large language model (LLM)-based generative AI tools have issues with plagiarism, AI hallucinations, and inaccurate or fabricated references. This raises legitimate concern about the utility, accuracy, and integrity of AI when used to write academic manuscripts. Currently, there is little clear guidance for healthcare simulation scholars outlining the ways that generative AI could be used to legitimately support the production of academic literature. In this paper, we discuss how widely available, LLM-powered generative AI tools (e.g. ChatGPT) can help in the academic writing process. We first explore how academic publishers are positioning the use of generative AI tools and then describe potential issues with using these tools in the academic writing process. Finally, we discuss three categories of specific ways generative AI tools can be used in an ethically sound manner and offer four key principles that can help guide researchers to produce high-quality research outputs with the highest of academic integrity.

AI snapshot

Written from the paper’s full text.

TL;DR

This paper offers ethical guidance for healthcare simulation scholars on using LLM-powered generative AI, especially ChatGPT, in academic writing. It reviews publisher policies and known risks, then proposes tiers of acceptable use and four author reflection questions.

Key findings

  • Publisher and editorial guidance converges on transparency, non-authorship of AI tools, and the primacy of authors’ creative and intellectual contributions; AI cannot be accountable for accuracy or approve manuscripts.
  • Cited studies report serious reference problems: among 50 ChatGPT-generated research protocols, 38% had a wrong or fabricated DOI and 16% of referenced articles were completely fabricated; among 30 short medical papers, nearly half of references were fabricated, 46% were authentic but inaccurate, and only 7% of all references were completely authentic and accurate.
  • The authors classify uses into ethical tiers: most acceptable uses restructure existing text (grammar/spelling, readability, translation); contingent uses include outlines, summaries, clarity edits, and brainstorming; suspect uses include drafting de novo text, data interpretation, literature review, and checking ethical compliance or plagiarism.
  • They propose four author checklist questions covering original ideas/interpretations, maintaining human competency, checking accuracy/reliability/bias of content and references, and disclosing exactly how generative AI was used.

Methods

  • This is a narrative/commentary paper, not an empirical study; the authors conducted a literature search and asked ChatGPT4, “What recommendations exist for ethical use of ChatGPT in academic writing?”
  • They reviewed the literature search and ChatGPT response, identified original source materials, discussed them, and incorporated edited content with original ideas from the authorship team.
  • They also asked ChatGPT 4.0 for a list of ways ChatGPT can be ethically used to assist authors writing for medical journals, then reflected on, modified, or excluded items to produce their suggested ethical-use framework.

Limitations

  • The paper states its concepts apply solely to ChatGPT and other similar LLM-powered generative AI tools, and that recommendations may need revision as technology evolves.
  • The paper states it did not discuss generative AI tools developed specifically for the academic community, such as Scopus AI, and calls for future research on those tools.
  • Evident from the text: the recommendations are based on literature review, ChatGPT outputs, and author reflection rather than systematic review or empirical validation, and the ethical tiers are presented as guidance rather than an exhaustive list of permissible or impermissible actions.

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

Integrity signals

Registry

Crossref · 10.1186/s41077-025-00350-6

Cites

18 of 18 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 48 cited sources checked · 0 supported · 0 flagged

  • Premisecitation

    LLM-based generative AI tools have issues with plagiarism.

    “Studies highlight three key challenges with content created by LLM-based generative AI tools: (a) plagiarism [5, 6];”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [5] Kim S-J. Research ethics and issues regarding the use of ChatGPT-like artificial intelligence platforms by authors and reviewers: a narrative review. Science Editing. 2024;11(2):96–106. source

      Not checked in the free run.

    • cites [6] Kacena MA, Plotkin LI, Fehrenbacher JC. The use of artificial intelligence in writing scientific review articles. Curr Osteoporos Rep. 2024;22(1):115–21. source

      Not checked in the free run.

  • Premisecitation

    Among references generated by ChatGPT for 50 research protocols, 38% had the wrong DOI or a fabricated DOI.

    “Amongst the references generated for these research protocols, 38% had the wrong DOI or a fabricated DOI”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [15] Athaluri SA, Manthena SV, Kesapragada V, Yarlagadda V, Dave T, Duddumpudi RTS. Exploring the boundaries of reality: investigating the phenomenon of artificial intelligence hallucination in scientific writing through ChatGPT references. Cureus. 2023;15(4): e37432.

      Not checked in the free run.

  • Premisecitation

    LLM-based generative AI tools have issues with inaccurate or fabricated references.

    “(c) inaccurate or fabricated references [13–17].”
    Quote in the paperSubstantial claim0 of 5 sources checked
    • cites [13] Alkaissi H, McFarlane SI. Artificial hallucinations in ChatGPT: implications in scientific writing. Cureus. 2023;15(2): e35179.

      Not checked in the free run.

    • cites [14] Buholayka M, Zouabi R, Tadinada A. The readiness of ChatGPT to write scientific case reports independently: a comparative evaluation between human and artificial intelligence. Cureus. 2023;15(5): e39386.

      Not checked in the free run.

    • cites [15] Athaluri SA, Manthena SV, Kesapragada V, Yarlagadda V, Dave T, Duddumpudi RTS. Exploring the boundaries of reality: investigating the phenomenon of artificial intelligence hallucination in scientific writing through ChatGPT references. Cureus. 2023;15(4): e37432.

      Not checked in the free run.

    • cites [16] Bhattacharyya M, Miller VM, Bhattacharyya D, Miller LE. High rates of fabricated and inaccurate references in ChatGPT-generated medical content. Cureus. 2023;15(5): e39238.

      Not checked in the free run.

    • cites [17] Benichou L, ChatGpt. The role of using ChatGPT AI in writing medical scientific articles. J Stomatol Oral Maxillofac Surg. 2023;124(5):101456. source

      Not checked in the free run.

  • Premisecitation

    Among references generated by ChatGPT for 50 research protocols, 16% of the referenced articles were completely fabricated.

    “and 16% of the referenced articles were completely fabricated”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [15] Athaluri SA, Manthena SV, Kesapragada V, Yarlagadda V, Dave T, Duddumpudi RTS. Exploring the boundaries of reality: investigating the phenomenon of artificial intelligence hallucination in scientific writing through ChatGPT references. Cureus. 2023;15(4): e37432.

      Not checked in the free run.

  • Premisecitation

    LLM-based generative AI tools have issues with fabricated or false information (AI hallucination).

    “(b) fabricated or false information (i.e. ‘AI hallucination’) [13–15];”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [13] Alkaissi H, McFarlane SI. Artificial hallucinations in ChatGPT: implications in scientific writing. Cureus. 2023;15(2): e35179.

      Not checked in the free run.

    • cites [14] Buholayka M, Zouabi R, Tadinada A. The readiness of ChatGPT to write scientific case reports independently: a comparative evaluation between human and artificial intelligence. Cureus. 2023;15(5): e39386.

      Not checked in the free run.

    • cites [15] Athaluri SA, Manthena SV, Kesapragada V, Yarlagadda V, Dave T, Duddumpudi RTS. Exploring the boundaries of reality: investigating the phenomenon of artificial intelligence hallucination in scientific writing through ChatGPT references. Cureus. 2023;15(4): e37432.

      Not checked in the free run.

  • Classified as probably not a substantial claim. 20
  • May combine two claims. 15

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

    Findingown data

    ChatGPT oversimplified many items in the list it provided, which may result in misleading guidance for academic authors and scholars.

    “We found that ChatGPT oversimplified many items in this list, which may results in misleading guidance for academic authors and scholars.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    Using ChatGPT for generating an outline, summarising content, improving clarity of drafted content, or brainstorming ideas tasks ChatGPT with generating novel text and thus has heightened potential for the introduction of bias, hallucination, or plagiarism if engaged in uncritically.

    “In each case, these uses task ChatGPT with generating novel text and thus have heightened potential for the introduction of bias, hallucination, or plagiarism if engaged in uncritically.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    The International Committee of Medical Journal Editors, JAMA Network Journals, and the World Association of Medical Editors have all issued specific guidelines on the use of generative AI.

    “Further, and especially relevant to the work of academic writers in healthcare simulation, the International Committee of Medical Journal Editors [10], JAMA Network Journals [7], and the World Association of Medical Editors [11] have all issued specific guidelines on the use of generative AI.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [10] Editors ICoMJ. Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals 2025. Available from: https://www.icmje.org/icmje-recommendations.pdf. Accessed April 4, 2025.

      Not checked in the free run.

    • cites [7] Flanagin A, Bibbins-Domingo K, Berkwits M, Christiansen SL. Nonhuman “Authors” and implications for the integrity of scientific publication and medical knowledge. JAMA. 2023;329(8):637–9. source

      Not checked in the free run.

    • cites [11] Zielinski C, Winker MA, Aggarwal R, Ferris LE, Heinemann M, Lapena JF Jr, et al. Chatbots, generative AI, and scholarly manuscripts: WAME recommendations on chatbots and generative artificial intelligence in relation to scholarly publications. Colomb Med (Cali). 2023;54(3): e1015868. source

      Not checked in the free run.

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

    Interpretationnone given

    LLM-powered generative AI tools can be used to provide translation of text into a semantically equivalent receptor language.

    “LLM-powered generative AI tools can also be used to provide translation of text into a semantically equivalent receptor language, which may be helpful for researchers looking to submit an article in a non-native language.”
    Quote in the paperNot a substantial claim

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

    Interpretationreasoning

    Generative AI tools cannot be held responsible for the accuracy of the manuscript's content, nor approve a submitted or published manuscript or vouch to support any subsequent investigation of claims made against the work.

    “These tools, however, cannot be held responsible for the accuracy of the manuscript’s content, nor approve a submitted or published manuscript or vouch to support any subsequent investigation of claims made against the work.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    ChatGPT is not designed to directly check grammar and spelling, but it will respond to prompts by rewriting content in a grammatically correct fashion without spelling mistakes.

    “While ChatGPT is not designed to directly check grammar and spelling, it will respond to prompts by rewriting content in a grammatically correct fashion without spelling mistakes.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Among the inaccurate references generated by ChatGPT, 48% were errors with the title, 52% represented errors with authorship, and 93% had the wrong PMID.

    “Amongst the inaccurate references, 48% were errors with the title, 52% represented errors with authorship, and 93% of these references had the wrong PMID.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [16] Bhattacharyya M, Miller VM, Bhattacharyya D, Miller LE. High rates of fabricated and inaccurate references in ChatGPT-generated medical content. Cureus. 2023;15(5): e39238.

      Not checked in the free run.

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

    Interpretationnone given

    Researchers ultimately hold full responsibility for the originality of their manuscript, accuracy and relevance of content, and appropriate referencing of published literature.

    “Researchers ultimately hold full responsibility for the originality of their manuscript, accuracy and relevance of content, and appropriate referencing of published literature.”
    Quote in the paperNot a substantial claim

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

    Interpretationreasoning

    The use of ChatGPT to perform the primary analysis of study data introduces the possibility of bias and hallucination and short-circuits an intellectual process that is most properly the domain of the authors.

    “The use of ChatGPT to perform the primary analysis of study data introduces the possibility of bias and hallucination and effectively short-circuits an intellectual process that is most properly the domain of the authors.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    The model is not explicitly utilising the meaning of tokens, but is rather generating syntactically appropriate strings of text based on statistical associations.

    “While the ‘meaning’ of these tokens is implicit to some extent within these associations, the model is not explicitly utilising this meaning, but is rather generating syntactically appropriate strings of text based on statistical associations [20].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [20] Ferrer J. How transformers work: a detailed exploration of transformer architecture 2024. Available from: https://www.datacamp.com/tutorial/how-transformers-work?utm_source=google&utm_medium=paid_search&utm_campaignid=19589720830&utm_adgroupid=157156377071&utm_device=m&utm_keyword=&utm_matchtype=&utm_network=g&utm_adpostion=&utm_creative=684592141199&utm_targetid=dsa-2218886984380&utm_loc_interest_ms=&utm_loc_physical_ms=9191512&utm_content=dsa~tofu~tutorial~machine-learning&accountid=9624585688&utm_campaign=230119_1-sea~dsa~tofu_2-b2c_3-nam_4-prc_5-na_6-na_7-le_8-pdsh-go_9-nb-e_10-na_11-na&ga

      Not checked in the free run.

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

    Premisecitation

    AI writing assistants (e.g. Grammarly) have been specifically designed for real-time writing support, proactive suggestions to rewrite sentences, and adjusting tone.

    “Authors looking for real-time writing support (e.g. grammar and spelling suggestions as you write), proactive suggestions to rewrite sentences, and adjusting the tone of writing should look to AI writing assistants (e.g. Grammarly), which have been specifically designed for these purposes [21].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [21] Grammarly vs. ChatGPT: choosing the right AI writing assistant. Available from: https://www.grammarly.com/blog/grammarly-vs-competition/grammarly-vs-chatgpt/#:~:text=Here's%20an%20example%20of%20Grammarly's,Writing%20Coach%2C%20to%20get%20feedback. Accessed December 1, 2024.

      Not checked in the free run.

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

    Interpretationreasoning

    It is possible that the issues of bias and hallucinations may be mitigated in academic generative AI tools.

    “it is possible that the issues of bias and hallucinations may be mitigated in these platforms.”
    Quote in the paperNot a substantial claim

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

    Interpretationreasoning

    ChatGPT will help with refining the readability and flow of the manuscript, assuring that it uses the best grammatical syntax.

    “It will also help with refining the readability and flow of the manuscript, assuring that it uses the best grammatical syntax.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    Generative AI tools are not designed to assess accuracy or authenticity of content, so they are prone to produce fabricated content when trained on large amounts of unsupervised data.

    “As generative AI tools are not designed to assess accuracy or authenticity of content, they are prone to produce fabricated content when trained on large amounts of unsupervised data, as is the case of ChatGPT.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    Transparency about the use of generative AI tools in the academic writing process is a cornerstone of academic integrity.

    “Firstly, and perhaps most importantly, transparency about the use of generative AI tools in the academic writing process is a cornerstone of academic integrity [7, 9, 11].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [7] Flanagin A, Bibbins-Domingo K, Berkwits M, Christiansen SL. Nonhuman “Authors” and implications for the integrity of scientific publication and medical knowledge. JAMA. 2023;329(8):637–9. source

      Not checked in the free run.

    • cites [9] COPE position - Authorship and AI - English. Available from: https://publicationethics.org/guidance/cope-position/authorship-and-ai-tools. Accessed December 1, 2024.

      Not checked in the free run.

    • cites [11] Zielinski C, Winker MA, Aggarwal R, Ferris LE, Heinemann M, Lapena JF Jr, et al. Chatbots, generative AI, and scholarly manuscripts: WAME recommendations on chatbots and generative artificial intelligence in relation to scholarly publications. Colomb Med (Cali). 2023;54(3): e1015868. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    There are no fact-checking components to this process in current models, which is an important source of AI hallucinations and propagation of biases within the training dataset.

    “There are no ‘fact-checking’ components to this process in current models, which is an important source of the AI hallucinations and propagation of biases within the training dataset.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationcitation

    These troublesome patterns may lead to the spread of misinformation or bias, or threaten researcher integrity if left unchecked.

    “These troublesome patterns may lead to the spread of misinformation or bias, or threaten researcher integrity if left unchecked [5].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [5] Kim S-J. Research ethics and issues regarding the use of ChatGPT-like artificial intelligence platforms by authors and reviewers: a narrative review. Science Editing. 2024;11(2):96–106. source

      Not checked in the free run.

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

    Interpretationreasoning

    Outsourcing some intellectual tasks to artificial tools runs the risk of necessary skills being lost over time.

    “Outsourcing some of these intellectual tasks to artificial tools runs the risk of these necessary skills being lost over time.”
    Quote in the paperNot a substantial claim

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

    Interpretationreasoning

    Even if authors double check content for accuracy, relevance, reasoning, bias, and appropriate referencing, they are still assuming some inherent risk given the issues described above.

    “Even if authors double check content for accuracy, relevance, reasoning, bias, and appropriate referencing, they are still assuming some inherent risk given the issues described above.”
    Quote in the paperNot a substantial claim

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

    Interpretationnone given

    ChatGPT is most skilled and accurate at text manipulation when the necessary ideas are already present in the initial query.

    “ChatGPT is most skilled and accurate at text manipulation when the necessary ideas are already present in the initial query.”
    Quote in the paperNot a substantial claim

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

    Interpretationreasoning

    Without appropriate content expertise, some authors may be unaware of AI-generated text that is plagiarised.

    “Without appropriate content expertise, some authors may be unaware of AI-generated text that is plagiarised.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Generative AI tools access publicly available data to generate text, which may produce content that closely or exactly resembles the original source of information.

    “Generative AI tools access publicly available data to generate text, which may produce content that closely (or exactly) resembles the original source of information [6].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [6] Kacena MA, Plotkin LI, Fehrenbacher JC. The use of artificial intelligence in writing scientific review articles. Curr Osteoporos Rep. 2024;22(1):115–21. source

      Not checked in the free run.

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

    Interpretationreasoning

    Drafting de novo text without providing original content in the prompt or developing new concepts for a particular section of text risks introducing factually incorrect material and deprives authors of an opportunity to engage deeply with the source material.

    “For example, drafting de novo text without providing original content in the prompt or developing new concepts for a particular section of text risks not only introducing factually incorrect material but also deprives the authors of an opportunity to engage deeply with the source material on which their paper is based on a personal level.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Interpretationcitation

    The data sources of academic generative AI tools are notably more reliable and robust.

    “Given that the data sources of academic generative AI tools are notably more reliable and robust, it is possible that the issues of bias and hallucinations may be mitigated in these platforms.”
    Quote in the paperSubstantial claim

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

    Interpretationreasoning

    Any literature search performed using ChatGPT must be confirmed with a traditional search engine, thus negating the actual benefit of using generative AI.

    “While some specialist products designed to support literature reviews are emerging (e.g. Elicit), any literature search performed using ChatGPT must thus be confirmed with a traditional search engine, thus negating the actual benefit of using generative AI.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    Mann et al. suggested three criteria for responsible use of LLMs in scholarship: human vetting and guaranteeing; substantial human contribution; and acknowledgement and transparency.

    “Mann et al. recently suggested three criteria for responsible use of LLMs in scholarship [23]: (1) Human vetting and guaranteeing; (2) Substantial human contribution; and (3) Acknowledgement and transparency.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [23] Mann SP, Vazirani AA, Mateo A, Earp BD, Minssen T, Cohen IG, et al. Guidelines for ethical use and acknowledgement of large language models in academic writing. Nature Machine Intelligence. 2024;6:1272–4. source

      Not checked in the free run.

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

    Interpretationreasoning

    Healthcare simulation scholars may not fully appreciate the scholarly benefits and scope of ethical risks associated with using LLM-powered generative AI tools for academic writing.

    “As a result, healthcare simulation scholars may not fully appreciate the scholarly benefits and scope of ethical risks associated with using LLM-powered generative AI tools for academic writing.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: There has been a notable paucity of research on this topic within the field of healthcare simulation.
27 of 27

  • The paper relies on this without support.

    There has been a notable paucity of research on this topic within the field of healthcare simulation.

    “There has been a notable paucity of research on this topic within the field of healthcare simulation.”

  • The paper relies on this without support.

    Some researchers have assiduously avoided engaging with generative AI due to a concern that they represent a threat to academic integrity.

    “While some researchers have embraced generative AI, others have assiduously avoided engaging with the tools due to a concern that they represent a threat to academic integrity.”
2 of 2

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?

Self-check

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

Each contradiction between two passages listed here was confirmed by an independent check.

  • Figure 1 identifies a different ChatGPT model

    Minor

    The text says the response shown in Fig. 1 came from ChatGPT 4.0, but the figure caption identifies it as a ChatGPT 4o response. These are different model designations.

1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. Scopus AI: Trusted content. Powered by responsible AI. Available from: https://www.elsevier.com/products/scopus/scopus-ai. Accessed December 1, 2024.

Cited by

Sorted by citations

Authors
Cheng, A., Calhoun, A., & Reedy, G.
Affiliations
  • University of Calgary (CA)
  • University of Louisville (US)
  • King's College London (GB)
Corresponding author
  • Cheng, Adam
  • Calhoun, Aaron
  • Reedy, Gabriel
ORCID
Venue
Advances in Simulation
Publisher
Springer Science and Business Media LLC
Published
2025-04-18
Volume
10
Issue
1
Pages
22-22
ISSN
2059-0628 (ISSN-L)
DOI
10.1186/s41077-025-00350-6
PMC
PMC12007126
OpenAlex
W4409588554
Semantic Scholar
5f22cb4379f76d3045e22fa53bee0c52d0d392ad
Cited by
140
References
18
Citations by year
2026: 99 · 2025: 41
FWCI
144.87
Type
journal article
Language
en
Open access
Licence
Locations
  • Advances in Simulation (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Libraries and Cultural Resources (University of Calgary) (submittedVersion)
Topics
  • Social Sciences › Safety Research › Academic integrity and plagiarism
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Social Sciences › Arts and Humanities › History and Philosophy of Science › Academic Writing and Publishing
Keywords
generative AI, academic writing, ChatGPT, large language models, AI ethics, academic integrity, plagiarism, AI hallucination, scholarly publishing, research ethics guidelines

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 2026-10-06.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-06.

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