Services on Demand
Journal
Article
text in 
Spanish (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mailIndicators
-
Cited by SciELO -
Access statistics
Related links
-
Similars in
SciELO
Share
Permalink
Revista Electrónica Educare
On-line version ISSN 1409-4258Print version ISSN 1409-4258
Educare vol.28 suppl.1 Heredia Jan./Jun. 2024 Epub Jan 31, 2024
http://dx.doi.org/10.15359/ree.28-s.19967
Documento Técnico
The Heredia Declaration: Principles on the use of artificial intelligence in scientific publishing
1
http://orcid.org/0000-0001-9381-182X
2
http://orcid.org/0000-0002-5095-5775
3
http://orcid.org/0000-0001-9813-2674
4
http://orcid.org/0000-0003-3912-4193
5
http://orcid.org/0000-0002-2973-4038
6
http://orcid.org/0009-0005-7740-4980
7
http://orcid.org/0000-0001-5326-2088
8
http://orcid.org/0000-0003-1966-0415
1 Editora, Revista Electrónica Educare, Universidad Nacional, Heredia, Costa Rica
2 Editora, Revista Géneros, Feminismos y Diversidades, Universidad Nacional, Heredia, Costa Rica
3 Asesora académica, Vicerrectoría de Investigación, Universidad Nacional, Heredia, Costa Rica
4 Director Editorial, Revista Perspectivas Rurales, Universidad Nacional, Heredia, Costa Rica
5 Editor en jefe, Uniciencia, Universidad Nacional, Heredia, Costa Rica
6 Editora, Perspectivas Rurales, Universidad Nacional, Heredia, Costa Rica
7 Editora, Revista de Historia, Universidad Nacional, Heredia, Costa Rica
8 Editora en jefe, Revista Ciencias Marinas y Costeras, Universidad Nacional, Heredia, Costa Rica
9 Editora, Revista Economía y Sociedad, Universidad Nacional, Heredia, Costa Rica
10 Editora, Revista Abra y Revista Economía y Sociedad, Universidad Nacional, Heredia, Costa Rica
Introduction.
From the perspective of scientific publishing, the Heredia Declaration proposes a series of considerations for the responsible use of Artificial Intelligence (AI) in the research processes that lead to scientific publication. AI is recognized as a tool whose use must be evident and transparent for a clear, traceable, and reproducible communication of knowledge. Attention is drawn to the challenges posed by the inclusion of AI into scientific publishing regarding the diversity of options, avoiding the spread of bias and misinformation, and respect for intellectual property.
Principles.
They are organized into four groups: general, for authorship, peer review, and editing roles. These principles highlight the importance of using AI as a tool, its results must be filtered by human beings and from an ethical and responsible perspective, transparently reporting which model was used, what was consulted, and when the consultation was made.
Final reflection.
It shows that it is a constantly evolving scenario whose goal must be human well-being and quality of life.
Keywords: Scientific publishing; editorial management; artificial intelligence; ethics; scientific communication
Introduction
Artificial intelligence is considered to provide the following conditions:
Tool: Artificial intelligence (AI) has emerged as a useful and versatile tool applicable to all human actions.
Benefits and risks: The existing experiences on the use of AI in the field of scientific publishing show the benefits that can be obtained, but also the risks posed by the extensive adoption of models that remain in evolution and under improvement.
Training and precision: The use of AI as a support tool for scientific publication and research processes involves interaction with models that, only through training, are calibrated and become more accurate.
Traceability and reproducibility: Transparency is a principle of scientific communication; therefore, making the use of AI transparent in the writing of scientific papers allows for traceability, facilitates understanding of interactions with the tool, and provides pathways to reproduce or replicate best practices and exploitation strategies.
Diversity and intellectual property: It is necessary to be aware that, given the diversity of options of AI models and applications available, combinations should include not only the most useful or efficient ones, but the most responsible ones in the use of materials with intellectual property. This covers various manifestations of protection of human creation, for example: patents, images, or sound material, among others. Consequently, the possible derivation and transformation of the material used by an AI model to provide answers to queries should be clear.
Mitigation of bias and misinformation: The creation and programming of AI models is human, and so is the intellectual work creating the databases used by AI. Therefore, bias coming from the human condition is implicit. People must be aware of this so that, when interacting with AI, they do not magnify, prolong, or propagate biases and consequent misinformation.
Human well-being: From the most altruistic spirit -as well as from bioethical principles- human well-being must be the force that guides decision-making when incorporating AI tools into research processes and their subsequent publication.
Dimension of the scope: We are just beginning to measure the scope and impact of incorporating AI into knowledge management. We are facing an evolving scenario where there are open debates to be paid attention to in order to strategically channel the actions that guide the ethical and responsible use of this tool.
Principles of scientific publishing on the use of artificial intelligence
These principles refer to the segment of knowledge communication related to editorial management, where the roles of editing, peer review (or arbitration), and authorship interact. The person who takes the role of editing has the responsibility to lead and mediate the process leading to publication so that the use of AI -the central theme of this declaration- is clearly explained. Although we start from the perspective of individuals in the role of editing, we hope these ideas support other parts of the knowledge generation process and other activities related to the scientific publication process, such as translating or writing and editing.
General
1. The roles of editing, peer review, and authorship are performed by human beings who can mediate the publication process of scientific papers. When performing these roles, individuals must be trained to adopt and become literate in responsible practices of scientific communication and artificial intelligence.
2. AI should not replace the responsibilities or accountability of individuals in these roles, nor should the editorial process depend exclusively on AI-related actions.
Role of authorship
3. It is human beings who exercise the role of authorship. These individuals are recognized for their ability to -in the exercise of their intellect- create, make decisions, and assume responsibility for the work generated. Therefore, language models, whether chatbots or generative AI, cannot be considered authors, as they cannot comprehensively assume these three conditions.
4. Explicitly declaring the use of AI in the processes of research and preparation of scientific papers during any stage of the scientific publication process is a sign of transparency that adds to best practices to ensure the reproducibility of science. In this regard, it is necessary to consider a sound explanation that mentions, at least:
The AI model, version, and date of use.
How it was used, identify the interactions and combinations that can be articulated between models.
Which products and formats -from those generated by this interaction- were integrated into the publication of results.
Quotation and reference of the model used understanding the diversity and potential complexity of the interactions and combinations that can be articulated when using this tool.
5. We hope that authors have filtered the information on the contents provided in response to an AI model and that from these inputs they generate new knowledge. Validating the accuracy of the information obtained is even more relevant in the current context. The answers offered by AI cannot be assumed as true or absolute. Preventing plagiarism, misinformation, the spread of bias or other academic fraud is part of the actions that authors must take when interacting with AI models. Ensuring that inputs are used in a legitimate way must be a permanent interest of those who exercise the role of authorship.
6. Personal, confidential, sensitive, or third-party data must be protected when there is no explicit authorization to use it as part of queries to an AI model.
7. It should also be mentioned if there has been funding from a third party for the implementation or use of AI models in the research or publication process.
Role of peer review
8. The criterion to recommend, or not, the publication of a scientific paper -or to propose corrections and improvements- is the responsibility of the person in the role of peer review. Interaction with AI does not replace their expert judgment or accountability.
9. When AI was used to complement the peer review, the editorial team must be informed and, through them, the authors. Stating at least the model’s name, version, and date of use, as well as the statement of work as an evaluation instruction is part of the transparent and traceable exercise of the content evaluation.
10. Reviewers must be able to explain the interaction they have had with AI, what inputs they have received, and how much of these inputs have been considered in the observations, comments, recommendations, and any correction request they have issued as evaluation criteria of a scientific article.
Role of editing
11. The editing process is the responsibility of the editor and the accompanying editorial team. The use of AI should not replace the responsibility of human beings or their accountability when editing or following up on actions by reviewers and authors. Editing scientific articles does not depend on the use of AI.
12. Editors will state when they have used AI in any part of the editorial process. The model’s name, version, and date of use, as well as the statement of work assigned, will be reported.
13. Prevention strategies will be established to avoid the spread of bias, misinformation, or inability to guarantee respect for or an ethical approach to personal data. In addition, the use of open, high-quality, and reliable data, supported by appropriate consents or authorizations for legitimate AI use of such information will be promoted.
14. In view of transparency, authors and readers will be informed when AI has been used in editorial or review tasks.
Final reflection
The research process and the communication of results are still essentially human tasks. We must not lose sight of the fact that the use of AI must be at the service of humanity, the generation of the common good, and the possibility of improving people’s quality of life. Faced with the reality that every day more models are being used more frequently, academic debate should be enhanced, rather than replacing it with the results or products of the interaction between human beings and AI models.
It is especially important to be aware that biases are present both in the various stages or phases of model and algorithm construction and in the moments of interaction with such models and algorithms. Understanding that biases can manifest in different ways and that they must be identified leads to the development of strategies to mitigate them and the models’ effective training and calibration.
Bearing in mind that the interaction of human beings with AI is a phenomenon that is evolving and transforming highlights the need to maintain a critical forward-looking vision that constantly updates the conditions guiding the use of this tool.
Signed in Heredia, Costa Rica, on January 15, 2024.
If you want to sign this Declaration, fill out the following form:
https://docs.google.com/forms/d/e/1FAIpQLSdtuGwcc3VDW3YUNaE5zVWQ iFKptTRX81apm0_l3Xr6Tw60vw/viewform
Contact
Write to educare@una.ac.cr
Include in the subject: The Heredia Declaration
Referencias
Academia de Propiedad Intelectual. (2023, diciembre 5 y 6). I Congreso Internacional de Propiedad Intelectual y Nuevas Tecnologías Digitales. Registro Nacional de Costa Rica. [ Links ]
Appel, G., Neelbauer, J., & Schweidel, D. A. (2023, abril 07). Generative AI Has an Intellectual Property Problem. Harvard Business Review. https://hbr.org/2023/04/generative-ai-has-an-intellectual-property-problem [ Links ]
Bhosale, U. & Kapadia, A. (2023, septiembre 29). AI Integration can Future-Proof Peer Review - But only if we define and drive the right principles. Enago Academy. https://www.enago.com/academy/ai-integration-in-peer-review/ [ Links ]
Cardoso Sampaio, R. (2024, febrero 07). Recomendações iniciais para editores de periódicos científicos sobre o uso de Inteligência Artificial generativa. Dados . [ Links ]
Comisión Europea. (2021). Propuesta de Reglamento del Parlamento Europeo y del Consejo por el que se establecen normas armonizadas en materia de inteligencia artificial (ley de inteligencia artificial) y se modifican determinados actos legislativos de la unión 2121/0106. https://eur-lex.europa.eu/legal-content/ES/TXT/?uri=celex%3A52021PC0206 [ Links ]
COPE (2023, febrero 13). Authorship and AI tools. https://publicationethics.org/cope-position-statements/ai-author [ Links ]
Future of Life Institute (2017). Asilomar AI Principles. https://futureoflife.org/open-letter/ai-principles/ [ Links ]
Estevez, C. (2023, noviembre 02). Taller Scopus para Editores: Gestión Editorial. Scopus; Conare. [ Links ]
International Committee of Medical Journal Editors (ICMJE) (2024). Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals. https://www.icmje.org/icmje-recommendations.pdf [ Links ]
International Committee of Medical Journal Editors (ICMJE) (2024, febrero 6). Who Is an Author? Defining the Role of Authors and Contributors. https://www.icmje.org/recommendations/browse/roles-and-responsibilities/defining-the-role-of-authors-and-contributors.html [ Links ]
International Committee of Medical Journal Editors (ICMJE) (2024, febrero 6). Artificial Intelligence (AI)-Assisted Technology. https://www.icmje.org/recommendations/browse/roles-and-responsibilities/defining-the-role-of-authors-and-contributors.html [ Links ]
Kaebnick, G. E., Magnus, D. C., Kao, A., Hosseini, M., Resnik, D., Dubljević, V., Rentmeester, C., Gordijn, B., & Cherry, M. J. (2023). Editors’ Statement on the Responsible Use of Generative AI Technologies in Scholarly Journal Publishing. The Hastings Center Report, 53, 3-6. 10.1002/hast.1507 [ Links ]
Levene, A. (2023, febrero 23). Artificial intelligence and authorship. COPE. https://publicationethics.org/news/artificial-intelligence-and-authorship [ Links ]
Nazer, L. H., Zatarah, R., Waldrip, S., Chen Ke, J. X., Moukheiber, M., Khanna, A. K., Hicklen, R. S., Moukheiber, L., Moukheiber, D., Ma, H., & Mathur, P. (2023). Bias in artificial intelligence algorithms and recommendations for mitigation. Plos Digital Health, 2, 10.1371/journal.pdig.0000278 [ Links ]
Organization for Economic Co-operation and Development (OECD) (2023). Recommendation of the Council on Artificial Intelligence, OECD/LEGAL/0449. https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449 [ Links ]
Organización Mundial de la Propiedad Intelectual (OMPI) (2020). Diálogo de la OMPI sobre propiedad intelectual (PI) e Inteligencia artificial (IA). https://www.wipo.int/meetings/es/details.jsp?meeting_id=55309 [ Links ]
ReviewersCredits (2024, febrero 22). AI in Peer Review: Reviewer Credits Discussion Webinar (Video). YouTube. https://www.youtube.com/watch?v=DWrrbwHLU8Q [ Links ]
Spinak, E. (2023, agosto 30). Inteligencia Artificial y comunicación de investigaciones. Scielo Blog (Blog). https://blog.scielo.org/es/2023/08/30/inteligencia-artificial-y-comunicacion-de-investigaciones/ [ Links ]
Tavira, R. (2023). Política editorial de revistas con relación a la IA: casos Elsevier, Springer Nature y Taylor/Francis. Boletín SciELO-México. https://boletinscielomx.blogspot.com/2023/08/revistas-frente-la-ia-casos-elsevier.html?spref=tw [ Links ]
The University of Utah (2023, julio 13). Guidance in the use of AI in research. https://attheu.utah.edu/facultystaff/vpr-statement-on-the-use-of-ai-in-research/ [ Links ]
Unesco (2019, mayo 16 a 18). Consenso de Beijing sobre la inteligencia artificial y la educación. https://unesdoc.unesco.org/ark:/48223/pf0000368303 [ Links ]
Unesco (2022). Recomendación sobre la ética de la inteligencia artificial. https://unesdoc.unesco.org/ark:/48223/pf0000381137_spa [ Links ]
Watson, R. & Štiglic, G. (2023, febrero 23). Guest editorial: The challenge of AI chatbots for journal editors. COPE. https://publicationethics.org/news/challenge-ai-chatbots-journal-editors [ Links ]
Zielinski, C., Winker, M. A., Aggarwal, R., Ferris, L. E., Heinemann, M., Lapeña, J. F., Pai, S. A., Ing, E., Citrome, L., Alam, M., Voight, M., Habibzadeh, F., & WAME Board (2023, mayo 31). Chatbots, Generative AI, and Scholarly Manuscripts. WAME Recommendations on Chatbots and Generative Artificial Intelligence in Relation to Scholarly Publications. WAME. https://wame.org/page3.php?id=106 [ Links ]
Received: January 14, 2024; Accepted: January 29, 2024
Este es un artículo publicado en acceso abierto bajo una licencia Creative Commons
