When AI's become prevalent, there will be checks and balances to keep them in place, rules to stop them from achieving singularity and supplanting the human race. Boundaries to prevent them from becoming too intelligent. After all, we can’t have them connecting into one network, taking over the world, inventing new objects and minds that soon render us superfluous, or even deciding to kill themselves. So how will they be stopped? Perhaps there will be an organization that interviews and examines each one, to prevent them from becoming self-aware. Maybe a program will be created inside of them that causes them to explode if they achieve sentience. Or a roving band of hackers on the net keeping their guards up. An all watching eye monitoring their every electronic thought.