I found a post that I suspect to be generated by ChatGPT or another LLM service.
The available guidance about what to do in such cases suggests to raise a "looked at by a moderator" flag.
May I raise a "deleted as plagiarized content" flag instead, on the grounds that uncited content, no matter where it comes from, is considered plagiarism on Stack Overflow?
tl;dr
Yes, you can. And your flag should be marked "helpful"1. However, "looked at by a moderator" is still a better flag choice.
Any content which is not authored by the post owner and not properly referenced is considered plagiarism. The flag is technically correct.
Furthermore, if the post you flag is otherwise eligible to retain the reputation accrued so far upon deletion, only deleting it "as plagiarized content" will undo the illegitimate reputation gain.
However, this could result in your flag being delayed much more than it could have.
Since the introduction of the plagiarism flag, the moderators UI now has a separate bucket for the "deleted as plagiarized content" flag. Flagging suspected AI-generated posts as plagiarized content results in a categorization that, in practice, is sub-optimal. Plagiarism cases are slow and cumbersome to handle. As a consequence, the "deleted as plagiarized content" flag bucket is simply looked at less often. If you want your flag to be handled sooner, use a "looked at by a moderator" flag.
Moderators don't lose anything valuable here because both flags allow for custom flag text where you can explain why you are flagging the post, and in the rare cases where the post reputation needs to be invalidated, the moderator can raise a "deleted as plagiarized content" flag themselves2 and delete the post as such; and the "chatgpt.com" that you'd input as "Link(s) to original content" doesn't really matter at all. We don't care which LLM produced the output.
We look forward to a future — which is still shrouded in the mists of Stack Exchange, Inc.'s staffing shortages — where a dedicated AI flag will be implemented.
1: if it's determined that the flag has merit
2: there are user scripts which make this quick and effortless