Plagiarism Policy

INTERACTION: Jurnal Pendidikan Bahasa

Originality and Responsible AI Use

Requirements for source attribution, similarity screening, and transparent use of artificial intelligence.

1. Originality and Attribution

Submissions must be original. Unattributed reuse of text, ideas, data, or images, including inappropriate self-reuse, is unacceptable. Identify quotations, cite accurately, and disclose substantial overlap with previous publications. Citation alone does not justify substantial unquoted copying.

2. Similarity Screening

Editors assess Turnitin or iThenticate reports using 20% as the initial similarity threshold. They examine matching sources, location, quotations, references, and standard methodological wording.

Below 20% does not guarantee acceptance; above 20% requires assessment, not automatic rejection. Similarity is not itself proof of plagiarism.

3. Assessment and Action

Authors may be asked to revise, clarify sources, or explain overlap and are given an opportunity to respond. Substantiated plagiarism may lead to rejection.

Post-publication findings may lead to correction or retraction, with notices linked to the article. Actions follow COPE Core Practices and Publication Ethics .

AI assistance does not transfer responsibility. Authors remain accountable for the originality, accuracy, integrity, and ethical compliance of their entire submission.

1. Permitted Assistance

AI tools may assist with language improvement, translation, organization, coding, or other clearly described tasks, subject to human oversight and appropriate disclosure. Their use must not replace the authors’ scholarly judgement.

Authors must check all outputs, verify references against original sources, and correct inaccuracies, bias, misleading statements, and unattributed material. AI-generated references must never be accepted without verification.

2. AI Cannot Be an Author

AI tools must not be listed as authors or co-authors. They cannot approve publication, accept accountability, or fulfil the journal’s Authorship Criteria .

3. Required Disclosure

Disclose generative AI used to generate or substantially rewrite text, translate passages, develop code, analyze data, or create other manuscript content. Identify the tool, provider, model or version where available, purpose, affected sections, and how outputs were checked.

Place a Declaration of AI Use before the References. During double-blind review, keep this declaration free of identifying author information.

Routine spelling, punctuation, and reference-formatting tools do not require disclosure unless their generative functions produce or substantially change content.

Example: Declaration of AI Use

During manuscript preparation, the authors used [tool, provider, and model/version] to [describe the purpose and affected sections]. The authors reviewed and revised the outputs, verified the relevant sources, and take full responsibility for the final manuscript.

Adapt this statement to the actual use. Do not include claims about checks that were not performed.

4. AI as a Research Method or Object of Study

When AI is part of the research, explain its role in the Method section. Report relevant tools, versions, access dates, prompts, settings, procedures, validation, and limitations sufficiently for readers to assess the work. Relevant prompts or outputs may be supplied as supplementary material, subject to privacy and copyright restrictions.

Synthetic data or AI-generated stimuli may be used when justified by the research design and clearly identified. They must never be presented as genuine participant responses, observations, or empirical evidence.

5. Unacceptable Uses

• Fabricating data, quotations, references, results, or participant accounts.
• Using AI to conceal plagiarism or evade similarity screening.
• Misrepresenting generated images, audio, or other content as authentic research records.
• Submitting unverified outputs or concealing AI use that requires disclosure.

6. Privacy and Confidentiality

Authors must not upload identifiable participant data, confidential records, or restricted material to external AI services without appropriate authorization, consent where required, and adequate data safeguards.

Editors and reviewers must not upload unpublished manuscripts or confidential review material to external generative AI services without editorial authorization and appropriate confidentiality safeguards. AI must not replace independent evaluation. Any authorized substantive assistance must be disclosed to the editor and, where relevant, the authors. Human editors retain responsibility for decisions.

7. Editorial Assessment

AI-detection scores, if used, are indicators requiring assessment, not proof of misconduct or grounds for automatic rejection. Similarity scores and AI-detection scores measure different things; the 20% similarity threshold is not an AI-use limit.

Editors may request clarification, supporting records, or a revised declaration and allow authors to respond. Substantiated breaches may lead to revision, rejection, correction, or retraction according to their seriousness and effect on the scholarly record.

Questions or Integrity Concerns?

Report concerns confidentially with article details, relevant source material, and supporting evidence.