What makes you think corporate hierarchies and processes won't infect their use of AI so that you won't be able to "go rogue" and you'll be held up by AI reviews, not humans.

Oh and the opposite when you are one of the 10 "other" people on someone else's project.