Hell humans already need AI protection from other AI. We need a web surfing companion that can easily mark accounts that are likely to be foreign influence farms and fury-generating bots. We need disinfo pointed out, and perhaps even help crafting our message so it comes across in the best way.