Ethics and Scientific Integrity in the use of Artificial Intelligence

Editorial Guidelines of the BJPE on the Ethical Use of Artificial Intelligence in Scientific Activity

Updated Version - July 2025

§1. Introduction

The use of artificial intelligence (AI) tools in the production, evaluation, and editing of academic texts has become increasingly common. While we recognize the potential of these technologies to improve the productivity and quality of work, it is essential to ensure that their use occurs responsibly, transparently, and with respect for the principles of scientific integrity. This guideline from the Brazilian Journal of Production Engineering (BJPE) establishes guidelines for authors, reviewers, and editors regarding the ethical use of generative and assistive AI tools within the journal. It is also important to emphasize that we always strive for compliance with the principles of COPE, Abepro, ABEC Brasil, ANPAD, ICMJE, IBICT, UFES, Dora, Barcelona Declaration, Hong Kong Principles, ABC, and UNESCO, reinforcing our commitment to scientific integrity, editorial ethics, and the responsible use of emerging technologies. Recognizing its role as a space for scientific training, BJPE understands that guiding, educating, and encouraging the ethical and critical use of AI tools is an essential part of its editorial mission.

§2. Author Guidelines

2.1. Mandatory Declaration. Authors must explicitly declare whether they used AI tools to assist in the production of the manuscript, specifying:
• which tool was used (e.g., ChatGPT, Copilot, Grammarly);
• for what purposes (e.g., grammar review, idea generation, translation, support in initial drafting, etc.); • the degree of subsequent human intervention.

This declaration must appear in the Acknowledgments section or in a footnote on the first page.

2.2. Prohibitions.

• AI tools cannot be listed as co-authors.

• The use of AI to automatically generate research results without human validation is prohibited.

• Submitting manuscripts generated entirely by AI, even with subsequent editing, is prohibited.

• The use of AI to manipulate data, simulate authorship, or fabricate bibliographic references constitutes scientific fraud and will be dealt with rigorously, potentially resulting in immediate rejection, public retraction, or communication to research institutions.

2.3. Authorial Responsibility. Responsibility for the manuscript's content rests entirely with the human authors. The use of AI does not exempt authors from obligations regarding originality, coherence, data accuracy, and compliance with BJPE standards.

§3. Guidelines for Reviewers

3.1. Permitted Use of Assistive AI. Reviewers may use AI tools to support the writing of reviews, provided that they: maintain critical and authorial judgment; fully review the generated texts; and do not include confidential manuscript data on open platforms.

3.2. Confidentiality. Manuscripts under review should not be shared with cloud-based AI systems that do not guarantee confidentiality and data protection.

§4. Guidelines for Editors

4.1. 4.1. AI Misuse Verification. Editors can use specialized tools (e.g., AI detectors, similarity checkers) to identify undeclared or inappropriate uses.

4.2. Adoption of Best Practices. Promote training campaigns on the ethical use of AI for all involved in the editorial workflow; encourage the declaration of tools used, even for text revision.

4.3. Institutional Alignment. This guideline is aligned with the recommendations of the aforementioned entities and reference bodies on integrity in scientific publications.

4.4. Warning against deepfakes. Generating images, citations, or graphics with AI that do not reflect factual reality (deepfakes) compromises scientific reliability and will be treated as misconduct.

§5. Practical Examples (Summary)

Use of AI to write the entire article - Prohibited

Translation with AI (reviewed by a human) - Must be declared

Use of AI to review grammar - Must be declared

§6. Glossary

Generative AI: creates original texts, codes, images, or music (e.g., ChatGPT, Copilot).

Assistive AI: corrects grammar, suggests words, translates (e.g., Grammarly, Deepl).

• Co-authorship: attributed only to those who contributed significantly with responsibility and the ability to account for the work.

§7. AI Usage Declaration Template

“We declare that we used the [AI name] tool for the purpose of [describe the function: spell checking, paragraph structure suggestions, translation, etc.], the text having been fully reviewed and validated by us, the human authors, who assume full responsibility for the content.”

This guideline will be periodically reviewed, keeping pace with the evolution of AI technologies and international editorial practices.

Prof. Dr. Rodrigo Randow de Freitas

Editor-Chefe – Brazilian Journal of Production Engineering (BJPE) Universidade Federal do Espírito Santo – UFES Centro Universitário Norte do Espírito Santo – CEUNES Departamento de Engenharias e Tecnologia – DET https://periodicos.ufes.br/bjpe/index Instagram: @revistabjpe