It would be cool if the more advanced AIs commented on you switching their morality. like if you have a tyrant AI that gets switched to protector he tells you that you should have chosen that a long time ago and stuff like that (them trying to stop you from changing them would also be cool but that would probably be harder to implement)
also it would be funny if the protector AI told you that you shouldn't have made it because it could have gone so wrong.