ABSTRACT
Background: Following the 2023 surge in popularity of large language models like ChatGPT, significant ethical discussions emerged regarding their role in academic authorship. Notable ethics organizations, including the ICMJE and COPE, alongside leading publishers, have instituted ethics clauses explicitly stating that such models do not meet the criteria for authorship due to accountability issues.Objective: This study aims to assess the prevalence and ethical implications of listing ChatGPT as an author on academic papers, in violation of existing ethical guidelines set by the ICMJE and COPE.Methods: We conducted a comprehensive review using databases such as Web of Science and Scopus to identify instances where ChatGPT was credited as an author, co-author, or group author.Results: Our search identified 14 papers featuring ChatGPT in such roles. In four of those papers, ChatGPT was listed as an “author” alongside the journal’s editor or editor-in-chief. Several of the ChatGPT-authored papers have accrued dozens, even hundreds of citations according to Scopus, Web of Science, and Google Scholar.Discussion: The inclusion of ChatGPT as an author on these papers raises critical questions about the definition of authorship and the accountability mechanisms in place for content produced by artificial intelligence. Despite the ethical guidelines, the widespread citation of these papers suggests a disconnect between ethical policy and academic practice.Conclusion: The findings suggest a need for corrective measures to address these discrepancies. Immediate review and amendment of the listed papers is advised, highlighting a significant oversight in the enforcement of ethical standards in academic publishing.
Disclosure statement
No potential conflicts of interest were reported by the author(s).
Author’s contributions
The first author assessed the databases. Other than that, the authors contributed equally to all aspects of the paper (conceptualization, literature search and analysis, writing and editing).
Notes
3. It is important to acknowledge that these recommendations possess an advisory nature, and editors retain autonomy in their decision-making processes, thereby maintaining the prerogative to adhere to or diverge from these recommendations.
4. Evidently, ChatGPT is not a group author, but is merely a coauthor. “A Group Author (sometimes called corporate author) is an organization or institution that is credited with authorship of an article by the source publication.” https://support.clarivate.com/ScientificandAcademicResearch/s/article/Web-of-Science-Core-Collection-Group-Author-field-definition-and-indexing-policy
5. We only found a stated “affiliation” of the US in six documents, while the rest did not list the affiliation, or simply stated “ChatGPT.”
6. At the ORCID registry, a search (https://orcid.org/orcid-search/search?searchQuery=ChatGPT) reveals two accounts with ChatGPT indicated as the “author” name: https://orcid.org/0009–0008–8990-591X; https://orcid.org/0009–0005–4505-850X
7. The precise reason that was given 22 March 2023 by the CMB EiC, Michael R. King, was “The opinions expressed in this manuscript do not warrant publication. Thus your manuscript will not be considered for publication in Cellular and Molecular Bioengineering.”