The people complaining about AI finding security vulnerabilities apparently prefer a world in which APTs and intelligence agencies continue freely infiltrating all of the world's companies, as they have been doing for decades.
Notably, those don't usually result in CVEs.