Why not? What's even the question to begin with? If a human is having trouble relating to another human, but society is bereft of human support to address that issue in an affordable manner, why should AI not be used? If AI is used to build a website coordinating some aspect of society like a community messageboard saying the pool is closed until the filter can be fixed, all of society must reject it?

Have you seen them select judges. They do filter for mundane and then select the crazies that typically hate women.

With Ai it will be about trust and whether it can optimize between wishes of masses, rights and work towards the best outcome.

Do I trust Ai? not off the box but if its pattern of past decisions are exposed, maybe. I don't hate Ai yet. I think people will corrupt them in time as they always do in future.