Policy on Artificial Intelligence and AI-Assisted Tools
AI-writing screening threshold: no higher than 20%. The journal has not yet selected a detection tool. Application of this numerical threshold will begin only after the tool and assessment procedure have been published and communicated to authors. Until then, no manuscript will be automatically rejected on an unspecified AI score. Detection scores are screening indicators, not a measurement of actual AI authorship or plagiarism; editorial assessment and an opportunity for authors to respond remain necessary.
When screening becomes operational, the editorial office will identify the tool, version, date, eligible text and report basis. Scores from different tools are not interchangeable. A score above the threshold prompts clarification and editorial review; a score below it does not establish originality or compliance. Authors may explain disputed results and request reassessment. Screening must respect manuscript confidentiality and applicable permissions.
AI systems and AI-assisted tools cannot be listed as authors because they cannot take responsibility for the work. Human authors remain fully responsible for originality, accuracy, citations, permissions, and all conclusions.
Authors must disclose substantive use of generative AI in preparing text, code, images, data analysis, or other research outputs. State the tool/provider and version where available, purpose, and which parts were affected, in an AI Use Declaration and, where relevant, the Methods section. Routine spelling and grammar checks do not require a declaration unless they substantially generate or alter content. Authors must verify every reference and factual claim; fabricated citations, data, results, or misleading images are prohibited.
AI used as a research method must be described sufficiently for assessment and reproducibility. AI-generated or materially altered images must be identified and justified; manipulation that misrepresents evidence is not permitted. Protect personal information, copyrighted material, and confidential data, and obtain necessary permissions before using external tools.
Reviewers and editors must not upload unpublished manuscripts, identifiable author information, review reports, or confidential editorial correspondence to public or third-party generative AI services. AI must not replace independent scholarly judgment or the editorial acceptance decision. Any permitted assistance must preserve confidentiality and be disclosed to the handling editor.
Example declaration: During preparation of this manuscript, the authors used [tool/version] for [purpose/sections]. The authors reviewed and verified the output and take full responsibility for the final content.
Suspected misuse will be assessed through the journal publication-ethics process, with an opportunity for authors to respond. Corrections or other proportionate editorial action may follow.
