Policy on the Use of Artificial Intelligence

The Editorial Board recognises that artificial-intelligence and automated tools can support research and scholarly communication. Their use must remain transparent, lawful, ethically appropriate, and subject to meaningful human oversight. The authors, reviewers, and editors remain fully accountable for their respective contributions and decisions.

This policy applies to generative AI, large language models, image and code generators, AI-assisted analytical tools, and other automated systems used in research, manuscript preparation, peer review, or editorial work.

1. Requirements for Authors

Disclosure

Authors must disclose the use of generative AI or other content-generating automated tools when their use goes beyond routine spelling, grammar, formatting, or reference-management assistance.

The disclosure should identify:

  • the name and version of the tool;
  • the purpose for which it was used;
  • the parts of the research or manuscript affected;
  • how the output was checked, corrected, and validated by the authors.

The disclosure should appear in the Methods, Acknowledgements, or a separate Declaration of AI Use section, as appropriate. Research in which an AI system is itself an object, method, or instrument of study must describe the system, data, parameters, prompts or procedures, validation, and limitations in sufficient detail to support critical assessment and, where feasible, reproducibility.

Authorship and Accountability

AI systems cannot be listed as authors or co-authors because they cannot take responsibility for a publication, approve the final version, declare conflicts of interest, or enter into copyright and licensing agreements.

Human authors are responsible for:

  • the accuracy, originality, and integrity of the manuscript;
  • verification of facts, calculations, code, analyses, images, and references;
  • identifying and correcting biased, misleading, unsafe, or fabricated output;
  • compliance with copyright, privacy, confidentiality, data-protection, and research-ethics requirements;
  • ensuring that all cited sources exist and support the claims made;
  • the final wording and scientific conclusions.

Generative AI must not be cited as an authoritative or primary source of scientific evidence.

Confidentiality and Rights

Authors must not upload confidential, personal, proprietary, security-sensitive, or third-party material to an AI service unless they have the right to do so and the service provides appropriate safeguards. Authors remain responsible for obtaining all necessary permissions and protecting research participants and affected organisations.

Unacceptable Uses

The following are unacceptable:

  • submitting AI-generated material without appropriate human verification and disclosure;
  • using AI to fabricate or falsify data, results, images, code, citations, or research procedures;
  • using AI-generated text or media that infringes copyright or contains unattributed material;
  • using AI to conceal plagiarism, duplicate publication, manipulation, or the absence of genuine research;
  • delegating the authors' essential intellectual responsibility or final scientific judgement to an AI system.

2. Requirements for Reviewers

Manuscripts and reviewer reports are confidential. Reviewers must not upload a manuscript, its figures, data, supplementary files, or unpublished ideas to public or third-party generative-AI systems.

Reviewers must not delegate the evaluation of a manuscript or the formulation of a reviewer recommendation to an AI system. The review must reflect the reviewer's own expert judgement. Any journal-approved automated assistance must preserve confidentiality, be transparently disclosed to the editor, and remain under the reviewer's direct control.

3. Requirements for Editors

Editors must not upload confidential submissions or reviewer reports to public or third-party generative-AI systems. They must not delegate assessment of scientific merit or editorial decision-making to AI.

Automated tools may support administrative checks, similarity screening, metadata validation, or integrity checks only when they have been appropriately assessed and are used under human supervision. Any concern identified by an automated tool must be reviewed by a qualified person before editorial action is taken.

4. Non-Compliance

Undisclosed or inappropriate use of AI may lead to a request for clarification or correction, rejection of the manuscript, notification of the authors' institution, or post-publication action, including correction, Expression of Concern, or retraction, depending on the seriousness of the case.

The policy will be reviewed periodically in response to changes in technology and scholarly-publishing standards. It is informed by relevant guidance from COPE and the European Association of Science Editors.