AI Usage Policy
Purpose
Generative AI and AI-assisted tools (for example, ChatGPT, Claude, Gemini, Copilot, DeepL, Grammarly, and similar tools) are changing how research is written and reviewed. JATI recognizes that these tools can support authors, but their use must be transparent, responsible, and under human control. This policy applies to authors, reviewers, and editors, and is consistent with COPE guidance on AI and authorship.
1. Policy for Authors
1.1 AI Cannot Be an Author
AI tools cannot be listed as authors or co-authors. Authorship carries accountability, consent, and responsibility for the work, which AI tools cannot take. Only humans can be authors.
1.2 Human Responsibility
Authors are fully responsible for the entire content of their manuscript, including any parts produced with AI assistance. This includes the accuracy of claims, the originality of the text, the validity of the analysis, and the correctness of every citation. Authors must carefully review and edit all AI-generated output.
1.3 Permitted Uses
The following uses are acceptable with disclosure where noted:
| Use | Disclosure Required? |
|---|---|
| Basic spelling, grammar, and punctuation checking | No |
| Language editing and improving readability of the author’s own text | Yes |
| Translation of the author’s own text | Yes |
| Generating or drafting text, summaries, or literature syntheses | Yes |
| Assistance with coding, data analysis, or qualitative coding | Yes, with a description in Methods |
| Generating images, figures, or illustrations | Yes, and subject to Section 1.4 |
| Brainstorming or outlining ideas | Yes |
1.4 Not Permitted
- Listing an AI tool as an author or citing it as a source of authority.
- Submitting unverified AI-generated content, including fabricated or inaccurate references, quotations, data, or results (“hallucinations”).
- Using AI to fabricate or falsify data, participant responses, interview transcripts, or findings.
- Using AI to create or alter images or figures in a way that misrepresents the underlying data. AI-generated images are allowed only for conceptual illustration, clearly labeled, and with confirmed rights to use.
- Entering participant-identifiable or confidential data into public AI tools in breach of consent or ethical approval.
- Using AI to disguise plagiarism, such as paraphrasing the work of others without attribution.
1.5 Disclosure Statement
Authors must include a statement in the Declarations section (before the References), and, for work that involves AI in the research methods, also in Methods. The statement should include the name and version of the tool, the date of use, and a description of how it was used.
Example: “During the preparation of this work, the authors used [tool name, version] to [purpose, such as improve the language of the Introduction]. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.”
If no AI was used, no statement is required, though authors may state: “The authors did not use generative AI tools in the preparation of this manuscript.”
2. Policy for Reviewers
- Reviewers must not upload a manuscript or any part of it, including the review report, to a public generative AI tool, because this may breach the confidentiality and proprietary rights of the authors.
- Reviewers must not use AI to produce the substance of their review. Peer review requires human expertise and judgment.
- Reviewers may use AI only for minor language polishing of their own review comments, provided no confidential content is uploaded.
- Reviewers who suspect undisclosed AI use in a manuscript should inform the editor.
3. Policy for Editors
- Editors must not upload submitted manuscripts or confidential correspondence to public generative AI tools.
- Editorial decisions are made by humans and are not delegated to AI tools.
- Editors may use secure, journal-approved tools for tasks such as similarity checking, plagiarism screening, and checking formatting, and these checks are always verified by a human.
4. AI in Research Design and Data
Where AI is itself part of the research (for example, a study of AI in learning, or AI used to analyze data), it is not governed by the disclosure statement alone. Authors must describe in Methods the tool, version, prompts or parameters (where relevant), the validation of its output, and the limitations and risks, and discuss ethical implications. Research on AI is welcome when it is explicitly examined as a mechanism, context, or outcome of transformative learning and social innovation (see Focus and Scope).
5. Detection and Consequences
JATI may use screening tools and editorial checks to detect undisclosed AI use, fabricated references, and other irregularities. Detection tools are imperfect and are never the sole basis for a decision; authors will be given the chance to respond. If a violation is found:
- Before publication: the manuscript may be returned, rejected, or the authors may be asked to provide clarification and corrections.
- After publication: the journal may publish a correction, an expression of concern, or a retraction, and may notify the authors’ institution, in line with the Publication Ethics policy and COPE guidance.
6. Policy Updates
AI technologies and publishing norms are developing quickly. JATI will review this policy regularly and update it as needed. Authors should check the current version before submitting. Questions can be sent to [email protected].
