Precursor of artificial intelligence regulation

Why is it that we always place regulation as a responsibility, most primarily, of governments? 

We think self-regulation has been of human's mode since the beginning of time, and entities working on advance technology or in the case of AIs, that without the usual apparent oversight, shall know their own work better and its effect and implications, if any, more than anybody else. 

Did we always say, honesty is the best policy? In AI self-regulation, whether it's done by the AI enterprise itself or within its industry, that will be definitively brooding.

What do you think? (We will update this shortly but you can comment now.)