It already does that, too, with the co-author

I would argue that is a net positive, it is valuable to know if a language model was involved enough to be committing itself.