Use of Generative Artificial Intelligence in Manuscript Preparation and Peer Review

Optica Publishing Group is monitoring the rapid developments and adoption of Large Language Models (LLMs) and Artificial Intelligence (AI) tools in scientific publishing. The following policy statements are provided to guide authors, reviewers, and editors.

Authorship. LLMs and AI tools (including generative AI chatbots) cannot accept the varied responsibilities required of authors (as outlined in the Obligations of Authors in the “Guidelines of the Optica Publishing Group Concerning Ethical Practices in the Publication of Research”) and therefore cannot be listed as authors on papers. 

Text and Image Integrity. Use of LLMs and AI tools to generate text is not permitted but they can be used to enhance and edit original text. Their use to generate figures, images, media, or code must be properly documented in the paper upon submission and must include the type of tool used (including the name, version, model, source) and a description of when and how the tool was used. Generative AI tools cannot be used to create, alter, or manipulate original research data and results such as plots or measurements. Any generative AI tools used to create figures from data, such as graphs or charts, should be listed in the figure caption.

The use of non-generative machine learning tools to manipulate, combine, or enhance existing text, images or figures should be properly documented in the relevant section of the manuscript upon submission.

Authors are responsible for reviewing and appropriately acknowledging content produced by these tools. As a condition of authorship, authors certify that all materials in the paper, including those produced by these tools, are accurate, because AI tools can generate authoritative-sounding output that can be incorrect, incomplete, or biased. Editors may decline to move forward with manuscripts if AI is used inappropriately.

Authors may use generative AI tools and AI-assisted technologies for editing and grammar improvement of their writing without disclosure. Authors should carefully review the results of such AI support.

AI Use by Reviewers and Editors. To protect the confidentiality of peer-reviewed materials, referees and editors should not upload the contents of submitted manuscripts, peer review reports, or decision letters into external AI-assistance tools, even if it is for the purpose of improving language and readability.