>I'm all for AI safety, but this sounds more like corpo preference safety to me.

the concept of safety can't really exist independently of who gets to decide what is and isn't safe