@simonzerafa @ct_bergstrom many do, but only some authors comply. That casts doubt on all submissions, unfortunately. And, at least for now, we can only really identify LLM use when authors submit with obvious editing failures. That's what this paper caught - also I document here as best I can:
peeraireview.com/search.html

The future is pretty simple... We have to view all papers starting in 2023 as potentially containing GenAI content, unless we know and trust the authors.