> They may not use AI to *check for vulnerabilities*

you didn't read the comment, did you?