I was submitting a revision to a ScholarOne platform for a SAGE journal yesterday. As part of the submission workthough required attention to a new checkmark box authors need to fill out. I wish I got a screenshot, so I could show the text exactly, but the basic idea was authors need to certify that they either did not include AI generated text or if they did use AI generated text, if the author had "disclosed" it. Two questions. One about ethics--supposing one did use AI in the process but the final work bears few if any marks of those revisions. Would this require a disclosure or would it be something akin to using a calculator or scratch paper to work out some math? Certainly, no journal would request an author disclose use of such a basic tool used in the process in the methods section. Sure, AI is a bit different than a calculator but my point is about the tool being one minor stage in a much larger process. Considering how many revisions authors make to small and large parts of manuscripts before submission, it seems like some AI text in some early draft (if AI were used) that did not carry through in the same format it would not be reportable. But I honestly don't know. Basically, would an author need to use a different disclosure for AI in process versus AI in final product? Are they held to the same standard? The other question is about writing procedures. Suppose an author did use AI and carried that text through to the final draft submission, in more or less the same form as the AI output it. Years ago, this would probably parallel the "debate" about how to cite a qualitative coding software (like NVivo or MAXQDA) or stats software (R or SPSS). Nowadays, it is common to do this and there are procedures in place, such as list the software, version, date of release, etc. But I don't know of any generalized standard in parallel for AI gen text. My field uses APA 7th, revised most recently in 2019, well before AI was as commonplace in academic writing. APA7 has no official guidance for this kind of thing (at least not that I know of--and nothing specific for AI, but maybe general for analysis software, per above). Maybe APA's blog has a suggestion, but if so, it would not yet be canon. And such, would be open to interpretation. How do authors do this? What would be an acceptable sentence an author might use to disclose AI in the text itself? Would it be like a citation and reference? A footnote or end note? Would one generalized comment cover the whole paper or would individual sentences or sections be cited/referenced? One statement, similar to COI disclosure or funding disclosure? Boilerplate text or more creative? Thanks, Profs. Happy Monday to y'all. We're on fall break for two days! How's by you?