AI Usage Policy

This journal uses artificial intelligence in a limited, disclosed and supervised way. This
page explains exactly where, so that authors and reviewers can make an informed judgement.

Where we use AI.
• Reviewer matching. When an editor looks for reviewers, the manuscript's title, abstract and
keywords, together with candidate reviewers' expertise profiles, are sent to Google Gemini,
which ranks candidates by topical fit. Candidates are filtered for conflicts of interest and
workload before any AI call, so the model never sees an ineligible reviewer.
• Editorial screening. The manuscript text may be summarised for the handling editor, with an
assessment of structure and language quality.
• Decision letters. An editor may request an AI-drafted letter synthesising the reviewer
reports. The editor edits and approves it before anything is sent.
• Reviewer assistance. Reviewers may optionally ask for private feedback on the tone and
completeness of their own comments. Their text is never altered.

What AI never does. The AI does not make editorial decisions. Every decision is taken by a
named human editor. An AI-generated-text likelihood score is advisory only and is never a
sole basis for rejection.

Author obligations. If you used generative AI in preparing your manuscript, say so in the
methods or acknowledgements. AI tools cannot be listed as authors: they cannot take
responsibility for the work.

Opting out. Authors who would prefer their manuscript not be sent to an external AI service
may say so in their cover letter, and we will handle it manually.