Peer review of a (clearly?) AI-generated paper: what to do?
I was recently asked to peer review a paper; I was very happy, it’s one of my first as an early career post doc, and the topic is of interest to me. However, I became increasingly disappointed with the sloppiness of the paper, starting from the writing (very long, convoluted sentences sometimes repeated verbatim one or two paragraphs later), to the analyses (entirely generated by Claude, not one tool was referenced or used and the figures are quite a giveaway) and even the references (some are duplicated, some are not fitting). There are a lot of numerical inconsistencies throughout the text and figures and there is no methodological transparency whatsoever. I’m very conflicted on what to do, straight up reject or suggest major revisions to try and save what’s worth saving? Is it true or just a myth that if a reviewer suggests rejection they just gets replaced until the paper gets eventually published? Thank you!!! Please be kind, I’m trying to learn myself 🙂