AIJOSH permits responsible use of generative AI while maintaining human accountability, research integrity, and confidentiality. This policy applies to authors, reviewers, and editors throughout submission, peer review, publication, and post-publication assessment.
1. Use by Authors
Authors may use AI to support literature exploration, idea development, organization, and language improvement. Such use must not replace their intellectual contribution, critical judgment, or responsibility for the work.
Authors must verify generated claims and references against reliable sources, check for bias, and protect confidential information and intellectual property. References suggested by AI must be checked against the actual publications before citation.
AI tools cannot qualify as authors because they cannot approve publication, accept responsibility, or meet the journal’s accountability requirements. Human authors remain responsible for the accuracy, originality, and integrity of all submitted content.
Substantive AI assistance must be disclosed at submission. This includes generating or substantially rewriting text, developing analytical content, interpreting research materials, or producing illustrations. Basic spelling, grammar, and punctuation checks, ordinary reference-management functions, and disability-related tools used solely for accessibility do not require disclosure.
AI use within research methods must be explained in the Method section, including the tool, version where available, purpose, procedures, and verification undertaken. Authors should provide sufficient detail to allow readers to assess the method and its limitations.
2. Research Data and Source Materials
AI-generated narratives, quotations, interview responses, archival references, reconstructed documents, or other synthetic materials must never be presented as authentic observations, participant testimony, or primary evidence.
Where synthetic data or AI-generated outputs are themselves part of the research design, authors must clearly identify them, explain how they were produced, and distinguish them from observed or collected data.
Authors using AI for transcription, translation, coding, or source interpretation must check outputs against the original materials. Uncertain readings, missing information, and interpretive limitations must remain identifiable. AI-generated additions must not silently fill gaps in the evidence.
For historical and documentary research, generated reconstructions must not be presented as original historical sources. Quotations, dates, names, and archival identifiers must be verified against the materials cited.
Interview recordings, personal information, unpublished research data, and restricted source materials must not be processed through external services without appropriate authorization and safeguards. Authors must respect participant consent, confidentiality commitments, intellectual property, and applicable access restrictions.
3. Figures and Images
AI assistance may be used for explanatory diagrams, timelines, and conceptual illustrations. Authors must check accuracy and disclose the tool, version where available, and purpose in the figure caption and AI declaration.
Data visualizations must faithfully represent actual research data through reproducible procedures. AI must not fabricate results, alter evidence misleadingly, or generate purported primary evidence.
General-purpose generative image tools must not be used to create graphical abstracts. AI-generated cover artwork requires prior approval from the editor and publisher.
Generated reconstructions must be clearly labeled as illustrations. They must not be presented as original photographs, maps, manuscripts, experimental images, or other research sources.
Where image generation or processing forms part of the research method, authors must describe the procedures and distinguish original materials from generated or modified outputs.
4. Disclosure Requirements
Place the declaration under Acknowledgments and Declarations, before the references, and provide matching information in the submission form where a corresponding field is available. Update the declaration if AI use changes during revision.
Disclosure must identify the tool, version where available, specific purpose, and extent of use. A general statement that “AI was used” is insufficient.
Authors may use the following template:
Use of Generative AI and AI-Assisted Technologies: [Tool and version, where available] assisted with [specific tasks or sections]. The authors checked the resulting material against [sources or verification procedures], revised it where necessary, and accept responsibility for the submitted work.
If no use requiring disclosure occurred, include:
Use of Generative AI and AI-Assisted Technologies: No generative AI use requiring disclosure under this policy occurred in preparing this manuscript.
The journal requires this negative declaration to support its submission checklist.
A general declaration does not replace methodological descriptions or figure-specific disclosures required elsewhere in the manuscript. Identifying details must be provided confidentially to the editor during anonymous review where necessary and restored before publication.
5. Use by Reviewers
Reviewers must not upload submitted manuscripts, extracts, supplementary files, or other confidential submission materials into AI tools.
Limited assistance with report language, organization, or general background searches is permitted only through tools with verified privacy safeguards and without exposing confidential content. Review reports may themselves contain confidential information and must be treated accordingly.
AI must not perform the reviewer’s independent scholarly assessment or determine their recommendations.
Reviewers must verify that inputs will not be reused, shared, used for training, or retained beyond the requirements of providing the service. Paid access alone does not establish privacy or authorize the processing of confidential information.
Reviewers must disclose permitted substantive AI assistance to the editor, identifying the tool and purpose in their report. Basic spelling and grammar checks are exempt. Reviewers remain accountable for every assessment, reference, and recommendation.
If compliance with these conditions cannot be established, the tool must not be used.
6. Use by Editors
Editors retain responsibility for manuscript assessment, reviewer selection, and publication decisions. AI must not replace editorial judgment.
Tools with verified privacy safeguards may assist with non-confidential administrative wording or general literature searches. Submitted manuscripts, extracts, reviewer reports, confidential correspondence, and other protected editorial records must not be uploaded to external AI tools.
AIJOSH does not treat access to a commercial AI service as authorization to process confidential submissions.
Any future journal-integrated AI service must undergo separate approval addressing confidentiality, data handling, security, and human oversight. Its intended use must be transparently communicated before implementation, and this policy must be updated where necessary.
7. Compliance and Policy Review
The journal may request clarification or supporting documentation about AI use. Concerns will be assessed under its Publication Ethics, Allegations of Misconduct, and Corrections and Retractions policies.
Editorial responses will depend on the available evidence, the nature of the concern, and its consequences for the reliability and integrity of the work. Actions may include clarification, additional disclosure, revision, rejection, or appropriate post-publication action.
Individuals concerned will have a reasonable opportunity to respond. Failure to disclose AI assistance will be assessed in context and will not automatically establish that the research findings are unreliable.
This policy will be reviewed periodically as technologies and publishing practices develop.
Source: Adapted from Elsevier’s Generative AI policies for journals, with journal-specific provisions for AIJOSH.