Artificial Intelligence (AI) and Automated Tools Policy

Document control: Version 1.0 | Effective date 21 October 2026 | Approved by Publisher | Next scheduled review October 2027 | First version.

AI tools may assist scholarly work, but they do not assume responsibility. Human authors, reviewers, editors, and publisher staff remain accountable for decisions, accuracy, originality, confidentiality, and ethical compliance.

Scope

This policy applies to generative AI, large language models, automated writing or editing systems, image generators, coding assistants, statistical or analytical automation, AI-based reviewer suggestions, similarity tools, image-integrity tools, and other automated systems used in research, manuscript preparation, peer review, editorial assessment, or production.

Use by Authors

• Authors must disclose substantive use of generative AI or automated tools in preparing the manuscript, analysis, code, figures, images, or other research outputs, beyond routine spelling, grammar, language correction, reference formatting, or similar low-risk assistance.

• The disclosure should identify the tool and describe how it was used in sufficient detail to permit editorial assessment.

• Disclosure shall be made in a dedicated section titled ‘Declaration of generative AI and AI-assisted technologies’ placed immediately before the reference list, and shall name the tool, provider and version, the purpose (e.g., language editing, code generation, literature summarisation) and the sections affected. Where no such tools were used, a statement to that effect is encouraged.

• Machine translation of a manuscript originally written in another language must be disclosed and the translation verified by the authors.

• Authors are responsible for verifying all AI-generated or AI-assisted text, data, code, citations, analyses, images, and claims.

• AI output must not be presented as evidence unless the research design specifically studies or validates that output.

• Authors remain responsible for obtaining necessary permissions and protecting confidential, personal, proprietary, or sensitive information.

Authorship

• AI systems, chatbots, and automated tools cannot be listed as authors or co-authors because they cannot take responsibility, approve publication, disclose conflicts, or enter legal agreements.

• Human authors must take full responsibility for all content submitted under their names.

Citations and Sources

• Generative AI should not be cited as a primary scholarly source for factual claims when an original verifiable source is available.

• Authors must verify references independently; fabricated, non-existent, or misattributed citations constitute a serious integrity concern.

• Where AI output is itself the object of study, the method of generation, model/version if known, date of access, and relevant prompts or parameters should be documented as appropriate.

AI-Generated or AI-Modified Images

• AI-generated or materially AI-modified scientific images, data visualizations, or evidentiary images must be disclosed and must not misrepresent underlying data.

• Image enhancement that changes scientific meaning, removes relevant features, invents features, or obscures manipulation is prohibited.

• Illustrative AI-generated artwork may be considered only when clearly labelled and legally permissible.

• Generative AI shall not be used to create, alter or enhance scientific data, images, figures, micrographs, blots or gels, except where the use of AI is itself the subject of the study and is fully described in the methods. Cover art and graphical abstracts generated by AI must be labelled.

• Studies that develop or evaluate AI models shall follow applicable reporting guidelines (e.g., TRIPOD+AI, CONSORT-AI, SPIRIT-AI, CLAIM, DECIDE-AI) and shall document model version, training data provenance, hyper-parameters, prompts and evaluation code sufficient for replication.

Use by Reviewers

• Reviewers must not upload confidential manuscripts, figures, data, supplementary files, or review correspondence into public or third-party generative AI systems without explicit editorial authorization and verified confidentiality protections.

• Reviewers should not delegate the intellectual task of peer review to generative AI. Review reports must reflect the reviewer’s own expert assessment.

• If a permitted tool is used only for language polishing of the reviewer’s own text, the reviewer remains responsible for confidentiality and accuracy.

• Reviewers shall declare in the review form whether any AI tool was used and for what purpose; undeclared or prohibited use may lead to removal from the reviewer database.

Use by Editors

• Editors must protect manuscript confidentiality and must not place unpublished content into external generative AI tools unless the publisher has approved the system and appropriate safeguards exist.

• Editorial decisions must be made by accountable human editors. Automated tools may support screening or triage but must not make final accept/reject decisions.

• AI-generated reviewer suggestions, integrity flags, or risk scores must be verified by a human before action is taken.

• Editors shall not use ‘AI-detection’ tools as the sole basis for any decision; such tools produce false positives and negatives and may be used only as one input to a human-led enquiry conducted under the Publication Ethics Policy.

Publisher Use of Automated Tools

Routine publisher use of automated systems for similarity checking, language assistance, metadata processing, reviewer discovery, figure analysis, or integrity screening should be disclosed where material. Such tools must be appropriately evaluated and used with human oversight.

Confidentiality, Privacy and Security

Users must consider whether an AI service stores prompts, uses submitted material for training, transfers data across jurisdictions, or exposes confidential content. Unpublished manuscripts and personal data must not be submitted to systems whose terms do not provide suitable protection.

Misuse and Editorial Action

Undisclosed or inappropriate AI use may result in requests for clarification or correction, rejection, withdrawal, correction, expression of concern, or retraction depending on severity and publication stage. Deliberate fabrication, falsification, deceptive image creation, fake citations, or concealment of material AI use may be treated as research misconduct. Where AI-generated content raises intellectual-property questions (e.g., unclear copyright status or inclusion of third-party material), authors must confirm that the content can be licensed under the journal's open-access licence.

Reference Framework

COPE Position Statement on Authorship and AI Tools (2023); ICMJE Recommendations, section on AI-assisted technology (current edition); WAME Recommendations on Chatbots and Generative AI in Scholarly Publications (rev. 2023); STM Generative AI in Scholarly Communications guidelines (2023); EQUATOR Network AI extensions (TRIPOD+AI, CONSORT-AI, SPIRIT-AI, CLAIM); Clarivate and Scopus research-integrity guidance.

Policy Review and Updates

Because AI tools and community guidance evolve rapidly, this policy shall be reviewed at least annually and whenever COPE, ICMJE or WAME issue updated guidance. The current version on the journal website prevails.