01 Scope and human responsibility

This policy covers generative AI used in preparing, assessing or communicating work submitted to Journal of Educational Innovation in Science and Technology. Authors remain responsible for every statement, reference, image and analysis in their submission. A tool cannot accept authorship responsibilities and must not be listed as an author. Human verification is required even when output appears fluent or technically convincing.

02 Disclosure in manuscript preparation

Disclose substantive assistance with the tool name, version where available, purpose, affected parts and the human checks performed. Place the declaration where editors and readers can find it. Routine spelling or grammar checks need not be declared under this policy. If no substantive assistance was used, authors should not insert a declaration that implies otherwise.

03 Research methods and visual material

When AI forms part of the research method, describe its role, relevant settings, validation and limitations in the methods section rather than only a writing declaration. Do not fabricate evidence, invent citations or present synthetic images as authentic observations. Clearly identify generated material where permitted and retain records sufficient to explain how it was produced and checked.

04 Privacy, rights and confidentiality

Before using a service, assess whether the input contains personal data, confidential research, third-party material or unpublished work that cannot lawfully be shared. Do not upload such information without appropriate authority and safeguards. A tool’s availability does not establish permission to use the data. Check outputs for inaccurate attribution, reproduced protected material and unintended disclosure.

05 Editors and reviewers

Editors and reviewers must not upload confidential submissions or reports to public generative AI services. Their assessment must remain an independent human evaluation. Any journal-authorised assistance requires prior safeguards and transparency about its role. Automated detection scores alone are not proof of inappropriate AI use; concerns require specific evidence and an opportunity for the author to respond.

06 Assessment and policy changes

Undisclosed or inappropriate use is assessed according to its significance, the explanation received and the reliability of the work. Possible responses include clarification, revision or the journal’s integrity and post-publication procedures. This policy is reviewed as tools and risks change. Material revisions require editorial review and approval before the revised policy is published.

07 Journal-specific provisions — retained content