Generative AI Policy — Journal of Social and Humanities (JSH)

JSH supports responsible, transparent use of Artificial Intelligence (AI) tools in research and scholarly writing. This policy clarifies what is permitted, what is prohibited, and what must be disclosed by authors, reviewers, and editors. It aligns with sector guidance from COPE and ICMJE and with emerging publisher norms.

1) Principles

• Human accountability: AI tools cannot be listed as authors; named authors are fully responsible for the content, including any text, data, code, and images prepared with AI assistance.
• Transparency: Any substantive AI use must be clearly disclosed in the manuscript.
• Integrity & privacy: Do not upload confidential or unpublished materials (manuscripts, reviewer reports, personal data) to public AI systems.

2) What Authors May Do

• Language editing and formatting: Use AI tools for grammar, spelling, and style polishing of your own text, with final human editing and full author responsibility.
• Idea structuring: Use AI to outline, summarize your own text, or generate alternative phrasings, provided you verify accuracy and remove fabricated claims or references.
• Code assistance (non-sensitive): Local/offline tools may be used to lint or refactor your own code if no confidential data/model weights are exposed and results are validated by the authors.

3) What Is Not Allowed

• Undisclosed AI-generated content: Submitting AI-generated text, images, tables, or translations without disclosure.
• Fabrication or falsification: Using AI to invent data, sources, quotes, or references (including “hallucinated” citations).
• Uploading confidential materials: Posting manuscripts, datasets with personal/identifiable information, or reviewer reports to public AI systems.
• AI as an author or co-author: AI tools cannot meet authorship criteria and must not be credited as authors.

4) Generative Images, Figures, and Visuals

Due to unresolved legal and research-integrity risks surrounding AI-generated images and videos, JSH does not permit AI-generated figures/images in the version of record unless the editors grant explicit, prior approval with documented rights and provenance. Data-derived figures must reflect actual analyses; illustrative graphics must not misrepresent findings.

5) Mandatory Disclosure (Authors)

Place a concise disclosure in the manuscript’s “Author’s Declaration” (before References). Example statements:

• “The authors used [Tool, Version] for grammar and language editing of author-written text. All content was reviewed and verified by the authors, who accept full responsibility for the work.”
• “The authors used [Tool, Version] to draft initial bullet-point outlines for the Introduction. All text in the submitted manuscript was subsequently written, checked, and approved by the authors.”
• “No generative AI tools were used in the writing, analysis, or figure preparation of this manuscript.”

If AI tools contributed to methods (e.g., translation pipeline, code generation, or model-assisted analysis), describe them in Methods with enough detail for replication. Do not cite AI tools as authors; instead, cite them as software/resources (with maker, model/version, and access date).

6) Reviewers and Editors

• Confidentiality first: Reviewers and editors must not upload manuscript content to public AI systems. Limited, local, privacy-preserving tools may be used for language polishing of the review text itself, never for processing the manuscript.
• Judgment remains human: AI may not be used to produce the review decision; it cannot replace expert evaluation.
• Disclosure: If an editor uses in-house screening tools (plagiarism, reference checks, basic language flags), this is handled within the journal workflow and does not expose author content externally.

7) Licensing and Attribution

JSH publishes under CC BY-SA 4.0. Any allowable AI-assisted contributions included in the article are released under CC BY-SA and may be adapted and shared under the same license terms. Authors are responsible for ensuring that any third-party content used by AI tools is license-compatible.

8) Detection, Verification, and Due Process

• Screening: The journal may use similarity and anomaly checks. Because AI detectors can be unreliable, editors will not rely on a single tool to make determinations.
• Clarifications: Where undisclosed AI use is suspected, editors may request explanations, underlying data/code, or an “AI usage log” (tool, purpose, location in manuscript).
• Remedies: Failure to disclose or misuse of AI may lead to correction, rejection, or post-publication action per the R–W–C Policy.

9) Relation to Other JSH Policies

• Peer Review Process — double-blind, confidentiality preserved.
• Open Access & Business Model — CC BY-SA; no fast-track; fees/penalties as stated.
• Author Guidelines — IEEE style; declarations; data availability.
• R–W–C Policy — actions for undisclosed AI use or integrity concerns.

10) Quick Reference

References & Useful Links

• COPE Position: Authorship & AI Tools
• ICMJE: Roles of Authors & Contributors (AI)
• Springer Nature: AI Editorial Policies (images & disclosure)
• Wiley: AI in Research Publishing