Declaration

Use of AI should be declared, minimally, in the Acknowledgements section. It may also be suitable to explain the use of such tools in the Methods section. However, the use of an LLM for copyediting purposes – e.g., to eliminate spelling and grammatical errors in human-generated text, or to make human-generated text more concise – does not need to be declared.

References

The discovery – during the review process – of AI-hallucinated references and/or citations will result in rejection of the paper without the possibility of reconsideration, unless the authors can show that the references in question were not, in fact, hallucinations. If the discovery is made after acceptance, during editing, copyediting or proof-correction, the paper will not be published. If the discovery is made after publication, the article will be retracted.

AI authorship

LLMs cannot be credited as authors as they cannot be held accountable for the work. Submission to this journal assumes that there is human accountability for every aspect of the submitted version of the text and that all authors agree that it is an accurate presentation of their work.

Generative AI images

Submissions should not include still or moving images produced by generative AI tools unless this is a necessary part of the work, in which case the use of generative AI must be declared and discussed with the editor. The use of non-generative image-editing tools, e.g., to combine still or moving images or to modify their brightness, tone or hue, is permitted unless the intention is to deceive.

AI use by peer reviewers

Peer reviewers play a vital role in the quality control of scientific publications. Their expertise is irreplaceable and their willingness to apply it to improve the work of others is very highly valued by editors. Peer reviewers are accountable for the accuracy and views expressed in their reports. In light of the confidentiality of the submission, the hallucinations that LLMs can produce, and the lawsuits brought by authors in response to the unauthorized use of their work to train LLMs, peer reviewers should not upload submissions into LLMs.