New Guidelines on the Use of Generative AI in Academic Writing
Effective October 2026.
To maintain the integrity and academic rigor of Thoughts, authors must disclose any use of generative AI in the research or writing process, including language editing, at the time of submission. This disclosure is mandatory in every submission, whether or not AI was used.
1. General PrinciplesGenerative AI (e.g., ChatGPT, Claude) may assist with tasks such as language editing, but it must not replace an author's intellectual contribution. Authors are fully responsible for verifying all AI-assisted content and take full accountability for the final manuscript. Findings, conclusions, and interpretations must be the author's own.
2. Disclosure Requirement (Mandatory)Every submission must include a disclosure statement, based on how AI was used:
- If AI was used for drafting manuscript text, generating images/graphics, or collecting/analyzing data → disclose in the Materials/Methods/Methodology section.
- If AI was used for anything else, including language editing/proofreading → disclose in a separate subsection titled "Statement on the Use of Generative AI and AI-Assisted Technologies," placed immediately after Acknowledgments.
- If AI was not used at all → use the same subsection above, with the "no AI" statement (see Section 3).
Definitions used in this section:
- Drafting: AI generates new text, or rewrites passages beyond correcting wording.
- Language editing: AI corrects grammar, spelling, punctuation, or phrasing while preserving the author's own wording and content.
- AI-powered writing assistants (e.g., Grammarly) count as AI use and must be disclosed. Basic spell-check built into word processors does not.
If AI was used, complete this template and place it in the required section (see Section 2):
The author(s) used [Name of AI Tool/Service and Version] for [General Purpose]. AI-generated output was used specifically for [describe how the output was incorporated and its scope]. After using this tool/service, the author(s) has/have thoroughly reviewed and edited the content and take(s) full responsibility for the content of the article.
Examples (for illustration only; describe your own use accurately and do not copy these statements):
Example 1 — language editing only (placed in the subsection after Acknowledgments)
Statement on the Use of Generative AI and AI-Assisted Technologies
The author used Claude Sonnet 5.5 (Anthropic, October 2026) for language editing and proofreading. AI-generated output was used specifically for improving sentence clarity and correcting grammar in the Discussion section; the argument, interpretations, and references are the author's own. After using this tool, the author thoroughly reviewed and edited the content and takes full responsibility for the content of the article.
Example 2 — several tools and purposes (placed in the subsection after Acknowledgments)
Statement on the Use of Generative AI and AI-Assisted Technologies
The authors used Grammarly Premium (v1.2.83) for grammar and punctuation checking, and Claude Sonnet 5.5 (Anthropic, October 2026) for translating three Japanese-language sources cited in the Literature Review for preliminary comprehension. AI-generated output was used specifically for correcting grammar and punctuation throughout the manuscript and for gaining an initial understanding of the three sources; the original texts were read in full, and all quotations and page references were checked against them. After using these tools, the authors thoroughly reviewed and edited the content and take full responsibility for the content of the article.
Example 3 — data analysis (placed in the Methods section)
Interview transcripts (n = 24) were first coded manually by the first author. Claude Sonnet 5.5 (Anthropic, October 2026) was then used to suggest candidate themes from the coded excerpts. Both authors checked each suggested theme against the transcripts, and only themes confirmed by both authors were retained. Transcripts were anonymized before use, and no identifying information was entered into the tool.
Fill in the brackets like this:
- Name of AI Tool/Service and Version — the tool, its developer, and the version or month of use, e.g., Grammarly Premium (v1.2.83); Claude Sonnet 5.5 (Anthropic, October 2026)
- General Purpose — e.g., language editing, translation for comprehension, bibliography formatting, brainstorming research angles
- Scope of use — e.g., improving clarity in the Discussion section only; translating three sources cited in the Literature Review for preliminary comprehension; formatting references in APA 7th edition
If AI was not used at all, include this statement instead:
The author(s) confirm(s) that no generative AI and AI-assisted technologies were used in any way in the preparation, writing, or completion of this manuscript.
Additional sample statements are available in the Faculty's Sample Statements on the Use of Artificial Intelligence [link/attachment].
4. Prohibited Uses- AI-generated content must not replace original analysis, argumentation, or scholarly insight.
- AI cannot be credited as an author — it cannot take responsibility for the work.
- Unverifiable AI-generated text, images, data, or references — including fabricated citations — must not be included in submissions.
- AI must not be used to process confidential or sensitive data.
- The journal may use AI-detection tools for preliminary screening. Results are indicative only and are not, on their own, grounds for rejection or a finding of misconduct.
- If a concern arises (e.g., a missing statement, a statement inconsistent with the manuscript, or unverifiable references), the editor will notify the author and give an opportunity to respond. The editor may request earlier drafts or file version history.
- Decisions rest on the manuscript's content, the author's response, and supporting evidence.
Depending on severity and the author's response, outcomes are:
- Revision: the author is asked to correct or complete the disclosure statement.
- Rejection: where AI replaced original analysis, or content cannot be verified.
- Retraction: where non-compliance is found after publication.
A false statement (for example, stating that no AI was used when it was) is treated as non-compliance.
