Thanks for this.
My first answer is that in focusing on governance and not use-cases, we are focused on shifting the grounds of power underlying this tech, and thus calling for different versions of the tech. This is different from the "AI for good" idea, which is about adopting the tools as they are for pro-social uses.