I'm not sure the experts have to be local. I can't be sure that a random twitter account isn't a bot, but I can be pretty sure that tweets from @nasa are reasonably trustworthy. People will form webs-of-trust: they trust one source, the people viewed as trustworthy by them, etc. Anyone outside of that will be untrustworthy.
That's not too dissimilar from what we do today, after all people have always been able to lie. The problem is just that if you start trusting one wrong person this quickly sucks you into a world of misinformation.
I find your point about regulating AI interesting. We already see some of this, with good recommendation systems being harmful to vulnerable people (and to a lesser degree most of us). This will probably explode once we get chatbots that can provide a strong personal connection, replacing real human relationships for people.
That's not too dissimilar from what we do today, after all people have always been able to lie. The problem is just that if you start trusting one wrong person this quickly sucks you into a world of misinformation.
I find your point about regulating AI interesting. We already see some of this, with good recommendation systems being harmful to vulnerable people (and to a lesser degree most of us). This will probably explode once we get chatbots that can provide a strong personal connection, replacing real human relationships for people.