They don’t have to hand over control to lose control to AI. People are easily manipulated, and AI has proven itself able to manipulate people. How long until a tech is tricked or coerced into doing something dumb on a planet scale, based on intentional misinformation given by its apparently benevolent AI assistant?