ON THE USE OF AI TOOLS
Recommendations for Authors
The use of tools and resources that assist authors in writing their manuscripts is acceptable, provided that the following principles followed by publications in the Netherlands are respected (see, for example, the site of the University of Leiden, referred below): honesty (reporting processes accurately and being open about margins of uncertainty), scrupulousness (being precise and thoughtful), transparency (showing one’s process and allowing others to build on it), independence (being impartial and unswayed by commercial or political interests), and responsibility (accepting accountability for the statements made).
Authors are free to use tools and resources that assist in the preparation, writing, revision, and translation of their articles, such as:
- Using AI to correct grammar, check spelling or make style improvements;
- Using AI to craft prose based a human-created draft or list of human-created ideas.
- Using AI to brainstorm ideas. But note that these ideas must be verified with actual sources by the human author.
- Using AI to modify images that are based on photographs or drawings that have been verified by the human author. You may wish to do this to remove extraneous items (e.g. backgrounds), to anonymize participants – especially images of children and identifiable people - or to provide a specific aesthetic effect.
In all cases, all uses of AI must be clearly identified and acknowledged.
Furthermore, JoSS reinforces that only humans can be authors, and in all cases, they must:
- Inform/mention the sources of the materials used. Any and all use or content generated by an AI application must be reported in the “AI Usage Statement” section at the end of the manuscript/final paper;
- Ensure that all cited material is correctly attributed, including complete citations, and that the cited sources support the statements of the AI application;
Some examples of unacceptable uses of AI are:
- Submission of prose, pictures, charts, tables, images created by AI that were not based on the creative product of the human author or a credited human source;
- Use of AI to misrepresent or obscure facts, even if unintentional;
Recommendations for Reviewers
Reviewers may use various tools provided by AI applications to check statistical analyses or references, for example. However, reviewers should not submit manuscripts to services that could cause the leakage of identities and content.
Although the journal has an AI usage policy for authors, reviewers may use resources that assist in detecting content generated or modified by AI tools. In this case, the use of these AI applications must be documented in the reviews. Concealing the use of AI tools is an ethical failure that violates the transparency of peer review.
Upon detecting any ethical failure in the manuscript, the reviewer must inform the editor.
How to cite AI content
Content generated by AI tools must be explicitly mentioned in the “AI Usage Statement” section of the manuscript/final paper. The non-use of any AI tool must also be mentioned in the same section.
Reference
University of Leiden integrity principles in research: https://www.staff.universiteitleiden.nl/research/quality-and-integrity/academic-integrity/academic-integrity.