There is nothing business-y about this. As a non-profit OpenAI can do whatever they want.
OpenAI isn't a single person, so decisions like firing the CEO have to be made somehow. I'm wondering about how that framework actually works.