SMAR’s Artificial Intelligence Use Policy sets out the principles governing the transparent, ethical, and responsible use of generative artificial intelligence and AI-assisted tools in academic publishing processes. Within this framework, the responsibilities of authors, reviewers, and editors, mandatory disclosure requirements, confidentiality principles, and procedures applicable in cases of policy violations are explained.
1. Use of Artificial Intelligence by Authors
SMAR adopts the principles of transparency, human accountability, and research integrity in the use of generative artificial intelligence and AI-assisted tools during the preparation and publication of scholarly studies.
Generative artificial intelligence or AI-assisted tools cannot be listed as authors or co-authors. Authorship requires accountability for the scientific content, accuracy, originality, and integrity of a study, and such responsibility can only be assumed by human authors.
Artificial intelligence tools must not be used as a substitute for the authors’ scientific judgment, critical evaluation, interpretation, or original scholarly contribution. Where authors use AI-assisted tools, they are responsible for verifying the accuracy, integrity, sources, originality, and ethical appropriateness of the content generated or suggested by such tools. Authors remain ultimately responsible for any incorrect, incomplete, misleading, or fabricated content generated by artificial intelligence tools.
Every manuscript submitted to SMAR must include an Artificial Intelligence Use Statement. Authors are required to state explicitly whether they have used generative artificial intelligence or AI-assisted tools for any purpose. All uses of AI-assisted tools, including but not limited to spelling, writing, grammar, punctuation, translation, formatting, text editing, rewriting, summarization, content generation, or similar purposes, must be disclosed.
Where an artificial intelligence or AI-assisted tool has been used, the statement must specify the name of the tool or service, the model and/or version where applicable, the purpose of use, and the scope of use. If no generative artificial intelligence or AI-assisted tool has been used during the preparation of the manuscript or the conduct of the research, this must also be stated explicitly.
Where artificial intelligence or AI-assisted tools are used in research design, code development, data generation, data processing, data analysis, coding, classification, modelling, or any other methodological stage of the research, such use must also be described in the Methods section in sufficient detail to enable scientific evaluation and, where applicable, reproducibility of the study.
Artificial intelligence tools must not be used to fabricate, alter, distort, or create non-existent research data, findings, results, references, or other scientific content.
Generative artificial intelligence or AI-assisted tools cannot be presented as authors or scholarly sources, and such tools cannot be cited as scientific sources. Authors are responsible for independently verifying the authenticity and accuracy of references, citations, data, and scientific claims generated or suggested by artificial intelligence tools.
Authors are responsible for protecting personal data, confidential information, intellectual property rights, and third-party content when using artificial intelligence tools.
2. Use of Artificial Intelligence by Reviewers
Manuscripts under peer review and all materials relating to their evaluation are confidential documents. Reviewers must not upload or share the whole or any part of an unpublished manuscript, including its figures, tables, supplementary files, or any confidential information relating to the manuscript, with third-party generative artificial intelligence systems.
The scientific evaluation contained in a peer review report must not be delegated to generative artificial intelligence or AI-assisted systems. Reviewer comments must reflect the reviewer’s own expertise, critical assessment, and academic judgment. Artificial intelligence tools cannot replace the reviewer’s scientific evaluation, and the reviewer remains fully responsible for the content of the submitted review report.
Provided that no confidential information relating to the manuscript or the peer review process is disclosed, reviewers may use AI-assisted tools solely to improve the language or expression of a review report that they have prepared themselves. Where such use occurs, the tool used and the purpose of use must be disclosed to the editor.
3. Use of Artificial Intelligence by Editors
Editors are responsible for protecting the confidentiality of manuscripts under evaluation, author information, reviewer identities and reports, editorial correspondence, and other confidential documents. Such content must not be uploaded to or shared with third-party generative artificial intelligence systems.
Scientific and editorial assessments concerning the scientific quality, originality, alignment with the journal’s aims and scope, methodological adequacy, or ethical status of a manuscript, as well as decisions concerning acceptance, revision, or rejection, must not be delegated to artificial intelligence systems. Artificial intelligence tools cannot replace the editor’s independent evaluation or ultimate decision-making responsibility.
Provided that no confidential manuscript- or peer-review-related information is disclosed, editors may use artificial intelligence tools for supportive purposes such as improving the language or expression of editorial correspondence. Nevertheless, the final editorial assessment, decision, and the content of the decision communicated to the authors remain the responsibility of the relevant editor.
4. Disclosure Requirement and Policy Violations
Failure to provide an artificial intelligence use statement, submission of an incomplete or misleading statement, or inaccurate disclosure of the AI tools used and the scope of their use will be evaluated within the framework of SMAR’s research and publication ethics principles.
The use of artificial intelligence tools to fabricate, alter, distort, or misleadingly generate data, references, citations, findings, analyses, images, or other scientific content may be considered a violation of research and publication ethics.
Results obtained from automated tools designed to detect artificial intelligence use are not considered, on their own, sufficient evidence of a policy violation. Suspected cases are subject to editorial review, and authors may be asked to provide an explanation, additional information, or supporting documentation where deemed necessary.
Where a use contrary to this policy is identified during the evaluation process, depending on the nature and seriousness of the violation, the authors may be asked to provide an explanation or make corrections, the evaluation process may be suspended, or the manuscript may be rejected. Where a violation is identified after publication, SMAR’s correction, expression of concern, or retraction procedures may be applied where appropriate.
Example Artificial Intelligence Use Statement
If artificial intelligence has been used:
“During the preparation of this study and/or the conduct of the research, the author(s) used [name of tool/service, model and/or version where applicable] for the purpose of [purpose of use] within the scope of [scope or section of use]. The outputs generated by the AI-assisted tool were reviewed, verified, and revised as necessary by the author(s). The authors assume full responsibility for the scientific content, accuracy, originality, and integrity of the study.”
If artificial intelligence has not been used:
“No generative artificial intelligence or AI-assisted tool was used during the preparation of this study or the conduct of the research.”