AI Use Policy

     RSDD – Editorial Policy on the Use of Generative Artificial Intelligence Tools – 2026

PRESENTATION

The use of Artificial Intelligence (AI) tools, especially those based on generative models, has increasingly become part of scientific production and the communication of knowledge. These technologies have been used to support activities such as language review, translation, information organization, programming, content synthesis, and other tasks related to the preparation of scientific manuscripts.

Recognizing this reality, the Revista de Segurança, Desenvolvimento e Defesa (RSDD) understands that the responsible use of these tools may contribute to the improvement of scientific communication, provided that the principles of scientific integrity, intellectual authorship, transparency, research ethics, and responsibility for the content produced are preserved.

At the same time, the use of Generative Artificial Intelligence poses challenges related to authorship, reliability of information, originality of manuscripts, confidentiality of the editorial process, and transparency regarding the resources employed during scientific production. These concerns appear in international guidelines on editorial ethics, especially regarding the impossibility of assigning authorship to AI tools, human responsibility for content, and the need to disclose the use of AI-assisted technologies (COPE, 2023; ICMJE, 2026; Elsevier, n.d.; Springer Nature, n.d.).

This Policy aims to guide authors, reviewers, editors, and the editorial team regarding the ethical, transparent, and responsible use of these technologies during all stages of the RSDD editorial process. The Policy was developed based on an analysis of national and international recommendations on scientific integrity, research ethics, AI governance, and good editorial practices, adapted to the characteristics and institutional objectives of the journal (CNPq, 2026; UNESCO, 2022; OECD, 2024).

IN SUMMARY

  • The RSDD does not prohibit the use of generative AI tools, but requires that their use be ethical, transparent, responsible, and compatible with scientific integrity.
  • Generative AI tools may not be listed as authors or replace human intellectual authorship.
  • Authors remain fully responsible for the content, sources, data, interpretations, and conclusions of the manuscript.
  • Any use of generative AI related to the preparation, review, translation, organization, or presentation of the manuscript must be clearly disclosed.
  • Reviewers and editors must preserve the confidentiality of the editorial process and must not enter manuscripts or reviews into external tools that may store, process, or reuse confidential content without institutional authorization.

1 Purpose and scope

This Policy establishes guidelines for the use of Generative Artificial Intelligence tools within the scope of the Revista de Segurança, Desenvolvimento e Defesa (RSDD), with the aim of promoting the ethical, transparent, and responsible use of these technologies in the process of submission, review, editing, and publication of scientific manuscripts.

Its purpose is not to restrict technological innovation or discourage the use of AI tools as support for academic and editorial activities. Rather, it seeks to establish parameters that ensure their use in a manner compatible with the editorial principles of the RSDD, scientific integrity, and the trust of the academic community.

This Policy applies to authors, coauthors, reviewers, editors, members of the editorial team, and other collaborators involved in the journal’s editorial process. Its provisions cover exclusively the use of generative AI tools related to the editorial activities of the RSDD. Non-generative automated tools, such as conventional spell checkers, reference managers, similarity checkers, and basic formatting resources, are not specifically addressed by this Policy, without prejudice to compliance with other applicable editorial standards.

2 Concepts and definitions

Artificial Intelligence (AI): an area of computer science dedicated to the development of systems capable of performing tasks normally associated with human intelligence, such as learning, pattern recognition, decision-making, and natural language processing.

Generative Artificial Intelligence (GenAI): a category of AI systems capable of generating, modifying, or transforming content, such as texts, images, computer code, tables, charts, audio, and videos, based on instructions provided by users.

Generative Artificial Intelligence tool: a computational application based on generative AI used to support academic, scientific, professional, or editorial activities. Examples include ChatGPT (OpenAI), Gemini and NotebookLM (Google), Claude (Anthropic), Microsoft Copilot, Perplexity AI, DeepSeek, among other tools with equivalent functionalities. This list is merely illustrative and does not limit the application of this Policy to specific tools.

Intellectual authorship: the human contribution responsible for the conception, development, analysis, interpretation, validation, and presentation of the scientific content of a manuscript.

Declaration of use of generative AI: information provided by the authors in the manuscript and/or during the submission process, indicating the tool used, its purpose, the stage at which it was used, and the human validation performed.

3 Guiding principles

This Policy is guided by principles that seek to balance technological innovation and scientific responsibility. These principles are aligned with international guidance on trustworthy AI, transparency, human oversight, and responsibility, as well as with editorial guidelines aimed at preserving authorship, integrity, and confidentiality in scientific publishing (UNESCO, 2022; OECD, 2024; COPE, 2023; ICMJE, 2026).

Scientific integrity: the use of generative AI must not compromise originality, data reliability, accuracy of sources, validity of results, or honesty in scientific reporting.

Transparency: the use of generative AI tools must be clearly reported, allowing editors, reviewers, and readers to understand how the technology was used.

Human responsibility: responsibility for submitted and published content always remains with human authors, reviewers, and editors, according to their respective roles.

Intellectual authorship: generative AI tools do not bear public responsibility for the content produced and, therefore, may not be listed as authors or coauthors of scientific manuscripts (COPE, 2023; ICMJE, 2026; ACM, 2026).

Editorial confidentiality: manuscripts, reviews, editorial communications, and unpublished data must be treated as confidential information.

Responsible innovation: the RSDD recognizes the potential of generative AI tools to support scientific communication, provided that their use is compatible with ethical and editorial standards.

4 Guidelines for author

The RSDD allows the use of generative AI tools as support for academic and editorial activities, especially when such tools contribute to the clarity, linguistic correctness, organization, and presentation of the manuscript. This use, however, does not replace the intellectual contribution of the authors, who remain responsible for all information, arguments, sources, data, interpretations, and conclusions presented in the work.

International recommendations converge on the understanding that AI tools should not be listed as authors, since they cannot assume public responsibility for the work, answer for its integrity, approve the final version of the manuscript, or account for possible errors (COPE, 2023; ICMJE, 2026; ACM, 2026).

4.1 Permitted uses, subject to disclosure

  • orthographic, grammatical, stylistic, or textual clarity review;
  • translation or support for the translation of excerpts of the manuscript;
  • organization of ideas, outlines, preliminary summaries, or auxiliary structuring of the text;
  • support for programming, technical data processing, corpus organization, or similar instrumental tasks;
  • support for the preparation of charts, tables, or syntheses, provided that the data and interpretations are checked and validated by the authors;
  • generation or support for the editing of images, charts, or visual materials, provided that the origin, purpose, and mode of use are explicitly disclosed.

In the cases above, AI must be understood as an auxiliary tool. Intellectual authorship, source selection, scientific reasoning, critical analysis, and final validation of the content remain human and non-delegable responsibilities.

4.2 Uses incompatible with the editorial policy

  • assigning authorship or coauthorship to generative AI tools;
  • submitting a text entirely produced by AI as if it were the intellectual work of the authors;
  • fabricating, altering, or manipulating data, results, citations, references, or evidence;
  • including nonexistent, inaccurate, or unverified bibliographic references;
  • using AI to replace scientific analysis, interpretation of results, or formulation of conclusions;
  • concealing the use of generative AI when its use has influenced the preparation, review, translation, organization, or presentation of the manuscript.

Special attention must be given to bibliographic references. International guidelines warn that AI tools may produce incorrect information, inaccurate citations, or nonexistent references; therefore, it is the authors’ responsibility to fully verify the existence, relevance, and accuracy of all cited sources (ICMJE, 2026; Elsevier, n.d.).

 4.3 Responsibility of authors

The use of generative AI tools does not reduce or transfer the authors’ responsibility for the manuscript. By submitting work to the RSDD, authors declare that they have checked, validated, and assume responsibility for all content presented, including passages reviewed, translated, organized, or suggested with AI support.

When generative AI has been used, authors must clearly inform the tool used, the purpose of use, the stage at which the tool was employed, and the manner in which the content was checked and validated. This requirement is aligned with practices adopted by scientific publishers that request a specific declaration of AI use in the manuscript or at the time of submission (Elsevier, n.d.; Wiley, n.d.; Taylor & Francis, n.d.; ICMJE, 2026).

5 Transparency and declaration of use

Transparency is the central axis of this Policy. Any use of a generative AI tool related to the preparation, review, translation, organization, analysis, presentation, or editing of the manuscript must be disclosed by the authors. The declaration is not punitive in nature; its purpose is to ensure clarity regarding the procedures adopted and to allow proper editorial evaluation of the work.

The declaration must appear in a specific section of the manuscript, preferably at the end of the text, before the references, under the title “Declaration on the Use of Generative Artificial Intelligence Tools.” When the journal’s submission system provides a specific field or checkbox, authors must also provide this information at the time of submission.

5.1 Minimum content of the declaration

  • name of the tool used and, when relevant, the provider or developer;
  • purpose of use, such as language review, translation, corpus organization, programming support, synthesis of information, or image generation;
  • stage of the work at which the tool was used;
  • indication that the content was checked, validated, and assumed by the authors;
  • any limitations or precautions adopted when the use of the tool involved data, images, code, or sensitive materials.

5.2 Declaration templates

Use for language review: The authors declare that they used [name of the tool] exclusively for language review, grammatical correction, and improvement of textual clarity. The scientific content, sources, analyses, interpretations, and conclusions were fully prepared, checked, and validated by the authors.

Use for translation: The authors declare that they used [name of the tool] to support the translation and/or review of the translation of excerpts of the manuscript. The final version was reviewed and validated by the authors, who assume full responsibility for the content presented.

No use: The authors declare that they did not use Generative Artificial Intelligence tools in the preparation, review, translation, organization, or presentation of this manuscript.

The RSDD may adjust the templates above in the Author Guidelines, in accordance with the evolution of good editorial practices and the requirements of research bodies.

6 Guidelines for reviewers

The peer review process is confidential. For this reason, reviewers must not enter manuscripts, excerpts from manuscripts, unpublished data, reviews, or editorial communications into external generative AI tools that may store, reuse, process, or train models with such information, unless there is express authorization from the journal and an institutional guarantee of confidentiality.

This precaution follows guidance from publishers and scientific organizations that warn of the risks of breaching confidentiality when manuscripts under review are submitted to external generative AI tools (Springer Nature, n.d.; Taylor & Francis, n.d.; ICMJE, 2026).

Reviewers may use AI tools for administrative support or language review of their own text, provided that they do not enter confidential content from the manuscript or editorial process. If the tool is used in a relevant way to support the preparation of the review, this use must be reported to the responsible editor.

Scientific evaluation, editorial recommendation, and the rationale of the review are the reviewer’s responsibilities. Generative AI tools must not replace critical judgment, methodological analysis, verification of argumentative coherence, or assessment of the manuscript’s scientific contribution.

 7 Guidelines for editors and the editorial team

 Editors and members of the editorial team may use generative AI tools to support administrative, linguistic, or organizational activities, such as reviewing editorial communications, standardizing texts, drafting documents, organizing workflows, and synthesizing non-confidential information.

Editorial use of AI must preserve the confidentiality of manuscripts, the decision-making independence of editors, and human responsibility for decisions made. AI may not replace editorial deliberation regarding admissibility, rejection, referral for review, acceptance, or publication of manuscripts.

Whenever the use of AI involves internal information, unpublished manuscripts, reviews, or sensitive data, the editorial team must observe the applicable criteria of confidentiality, information security, and institutional authorization. The use of public or external tools to process confidential content may occur only if there is express authorization from the journal and adequate assurance of data protection.

In cases of suspected undeclared or inappropriate use of generative AI, editors must prioritize careful, proportionate, and well-founded inquiry, avoiding decisions based exclusively on automatic AI detectors, whose results may present technical limitations and a risk of false positives.

8 Editorial procedures in cases of noncompliance

Noncompliance with this Policy will be analyzed according to the nature, severity, and effects of the observed conduct. Editorial action must observe criteria of proportionality, procedural transparency, and preservation of scientific integrity, in dialogue with the logic of education, prevention, investigation, and accountability present in national scientific integrity policies (CNPq, 2026).

Depending on the case, the RSDD may adopt one or more of the following measures:

  • request clarification from the authors;
  • require the inclusion, correction, or supplementation of the declaration of AI use;
  • return the manuscript for adjustments before proceeding with the review;
  • temporarily suspend editorial processing until the facts are investigated;
  • reject the manuscript when there is evidence of use incompatible with scientific integrity;
  • adopt post-publication measures, such as correction, expression of concern, or retraction, when the problem is identified after publication.

Formal omission of the declaration, when not accompanied by evidence of bad faith or harm to scientific integrity, may initially be addressed through a request for correction. Situations involving data fabrication, nonexistent references, manipulation of results, improper appropriation of authorship, or deliberate concealment of relevant use may give rise to more serious editorial measures.

9 Policy update

This Policy must be periodically reviewed by the RSDD Editorial Committee, considering the evolution of AI technologies, national standards, international recommendations, and good practices in scientific publishing.

Future updates may adjust definitions, examples, declaration templates, editorial procedures, and submission requirements, while preserving the principles of scientific integrity, transparency, and human responsibility.

REFERENCES

ACM. ACM Policy on Authorship. New York: Association for Computing Machinery, 2026. Disponível em: https://www.acm.org/publications/policies/new-acm-policy-on-authorship. Acesso em: 25 jun. 2026.

CNPQ. Portaria CNPq nº 2.664, de 6 de março de 2026: institui a Política de Integridade na Atividade Científica do CNPq. Brasília: Conselho Nacional de Desenvolvimento Científico e Tecnológico, 2026. Disponível em: https://www.gov.br/cnpq/pt-br/assuntos/noticias/cnpq-em-acao/cnpq-publica-portaria-que-institui-politica-de-integridade-na-atividade-cientifica. Acesso em: 25 jun. 2026.

COPE. Authorship and AI tools. Committee on Publication Ethics, 2023. Disponível em: https://publicationethics.org/guidance/cope-position/authorship-and-ai-tools. Acesso em: 25 jun. 2026.

ELSEVIER. Generative AI policies for journals. Amsterdam: Elsevier, [s.d.]. Disponível em: https://www.elsevier.com/about/policies-and-standards/generative-ai-policies-for-journals. Acesso em: 25 jun. 2026.

ICMJE. Use of Artificial Intelligence in Publishing. International Committee of Medical Journal Editors, 2026. Disponível em: https://www.icmje.org/recommendations/browse/artificial-intelligence/. Acesso em: 25 jun. 2026.

OECD. Recommendation of the Council on Artificial Intelligence. Paris: Organisation for Economic Co-operation and Development, 2019, atualizada em 2024. Disponível em: https://legalinstruments.oecd.org/en/instruments/oecd-legal-0449. Acesso em: 25 jun. 2026.

SPRINGER NATURE. Editorial policies. London: Springer Nature, [s.d.]. Disponível em: https://www.springernature.com/gp/policies/editorial-policies. Acesso em: 25 jun. 2026.

TAYLOR & FRANCIS. AI Policy. London: Taylor & Francis Group, [s.d.]. Disponível em: https://taylorandfrancis.com/our-policies/ai-policy/. Acesso em: 25 jun. 2026.

UNESCO. Recommendation on the Ethics of Artificial Intelligence. Paris: UNESCO, 2022. Disponível em: https://www.unesco.org/en/artificial-intelligence/recommendation-ethics. Acesso em: 25 jun. 2026.

WILEY. AI guidelines for researchers. Hoboken: Wiley, [s.d.]. Disponível em: https://www.wiley.com/en-nl/publish/article/ai-guidelines/. Acesso em: 25 jun. 2026.