Citing Generative AI

Specific journals and publishers may have their own guidelines and publication rules about use of generative AI, both text- and image-based. 

We will collate some of these guidelines here, but you should always double check with the journal or publisher to see their up-to-date information regarding generative AI use. 

Science

AI-assisted technologies, such as large language models, generative AI, and generative AI image creators cannot be listed as authors or coauthors, and sources that are authored or co-authored by such AI tools are not eligible for use in a Science published paper. If a researcher wants to use generative AI tools in their research, they must:

• note use in the cover letter and acknowledgements sections
• provide detailed information in the methods section, including: the full prompt used, the AI tool and version
• account for all accuracy and plagiarism potential by thoroughly checking the output of the AI tool
• ensure that all sources are appropriately cited 
• guard against bias that might be introduced by AI

AI-generated images and other multimedia such as charts and graphs are not permitted in Science journals without explicit permission from the editors (with exceptions based on research context, such as research about generative AI tools)

These policies may change, and Science will update this guide as needed.

Nature

Large language models and text-based generative AI tools should not be listed as authors or coauthors. If you use a generative AI or LLM tool you must properly document the use in the Methods section (or suitable alternative) in your manuscript. 

Generative AI images are not allowed in Nature, as published by Springer. Exceptions can be made for images from agencies with which Nature or Springer has contractual relationships, and images and video related directly to research about AI (which will be reviewed on a case-by-case basis).

Peer reviewers are not allowed to upload manuscripts into any AI tools.