Lorenzo García Aretio
Director/Editor of RIED
José Luis García Boyé
Technical Secretary of RIED
Throughout the past year, the debate concerning the impact of artificial intelligence (AI) on scientific publishing has intensified across various forums. From RIED—Revista Iberoamericana de Educación a Distancia—we wish to express our perspective, one that aligns on numerous points and diverges on others from equally respectable positions circulating publicly. This issue often concerns many editors of scientific journals, who are aware of their responsibility to uphold rigorous standards and quality requirements in publishing research.
The True Focus of Scientific Journals
We believe the primary objective of a scientific journal is not to evaluate the capacity, intelligence, or professionalism of an author but rather to disseminate knowledge and, above all, contribute to the construction and advancement of the specific scientific field addressed by the publication. It is not about assessing whether a text was exclusively authored by a human being, but about ensuring articles offer valuable contributions to knowledge and maintain academic rigor. In this regard, AI—like other tools (spreadsheets, statistical software, translators, etc.)—can enhance clarity and accessibility without diminishing authors' merit.
Common Objections and Our Rebuttal
- “AI can generate plagiarism or nonsensical texts.”
The issue lies not with the tool but with the content itself. Plagiarism and nonsensical texts have always existed, whether produced by a human or AI. While addressing plagiarism remains essential, the core focus should be the rigor, novelty, and scientific contributions of the article, rather than the method of its creation. - “Relying on AI to write an article demonstrates a lack of fundamental competencies.”
This assertion is, in our opinion, largely irrelevant. The role of scientific journals is to publish original content that advances knowledge, not to assess authors’ writing skills. Researchers' primary role is generating innovation, and AI can serve as a vital instrument for presenting these contributions more clearly and accessibly, thus enhancing the value of the content. Currently, a true deficiency in basic competencies among authors publishing in prestigious journals would be the inability or unwillingness to utilize available AI tools. - “Reviewers should not use AI.”
This claim appears inconsistent and of limited practical value. If AI can enhance and expedite the review process, prohibiting its use only hampers efficiency. It would be wiser to establish guidelines allowing reviewers to combine their expertise and judgment with AI support whenever appropriate. - “Plagiarism detection tools can indicate excessive AI usage.”
Currently, these systems are unreliable and prone to significant error rates, sometimes mislabeling human-written texts as AI-generated. Using them as rejection criteria poses risks to editorial integrity. Although these tools will improve over time, accurately determining which parts of a text were AI-assisted will become increasingly difficult.
Transparent Use and Authorial Responsibility
- Declaration of AI Usage: Authors should be required to transparently specify AI usage within the methodology, similarly to how other scientific tools are reported. They should describe clearly why, how, and in which sections of the work AI was employed. There is neither reason to hide nor penalize this practice.
- Authors' Responsibility: Human authors remain fully accountable for the texts they submit, irrespective of the tools used. This point must be explicitly clarified throughout review and publication processes.
- Ethics and Data Verification: Utilizing AI does not absolve authors of their responsibility to verify the accuracy of information, coherence of arguments, and proper citation of sources. Transparency in data collection, analysis, and interpretation remains an essential requirement in academic publishing.
- Training and Digital Literacy: Continuous education for authors and reviewers is recommended to effectively and ethically leverage AI technologies. Adequate digital literacy will facilitate the responsible integration of these tools, thereby enhancing the quality and rigor of scientific research.
- Reducing Linguistic Barriers: AI can help overcome language obstacles, especially benefiting researchers less proficient in the journal's primary language. This potential increases author participation and encourages the internationalization of knowledge.
Perhaps the Primary Focus of the Debate is Misplaced
- Constant Evolution of AI: AI advances so rapidly that rigid regulations quickly become obsolete. It is virtually certain that tomorrow’s circumstances will differ significantly.
- Near-Complete AI Production: What is currently perceived as imperfect AI-generated content may soon achieve considerable precision. It is highly probable that we will witness articles almost entirely conceived, researched, and written by AI; what will truly matter is whether these articles generate new knowledge.
- Current Competency: Presently, lacking proficiency in AI represents a clear deficiency in academic competencies.
Conclusion: AI is not the Problem; it is Part of Evolution
- Beyond the Mode of Writing: Excessive attention should not be given to who or how an article is written, but rather to the scientific contribution it provides.
- Tools for Clarity: Any resource that facilitates the clear and rigorous generation and expression of ideas is beneficial. We should not penalize methods that enhance our research.
- The Present, Not the Future: AI is not merely the future; it is already the present. Adapting to this reality is academia's only viable option.
- AI as an Ally: Rather than posing a threat, AI can become a valuable ally in enhancing publication quality, streamlining review processes, and reducing linguistic barriers, provided it is implemented transparently and responsibly.
- A New Paradigm: In this new scenario, ethics, continuous training, and responsible AI usage must become naturally integrated within research practice. Only through this approach can science evolve, democratize, and foster the generation of robust, open, and accessible knowledge for the entire academic community.
RIED
We take this opportunity to highlight that, several months ago, RIED included the following statement in its guidelines:
The use of artificial intelligence tools is permitted, provided authors clearly specify the type of AI used, the specific sections involved, and the purpose. This must be indicated in the methodology section or corresponding part of the paper. Using AI does not exempt authors from full responsibility for content, nor should it affect the authenticity or scientific integrity of their work.
Note: This entry, in its final version, was supported by AI.
