@elilla biorxiv (preprint server for biology) tried that two years ago.
Authors disagreed, as the abstracts were very low-quality, misrepresenting the science and usually failing at the task of offering a summary understandable for a general audience. They added a notice indicating that they were written by an AI tool and not approved by the authors.
Haven't seen any of them recently (good news!).
https://www.nature.com/articles/d41586-023-03545-x (behind paywall, can share pdf if you want it)