I could see the possibility of an AI forwarding those difficult and complex situations to a real human or offering that as a choice. I was reading about Minecraft the other day and did not realize this but Microsoft employs over half of the Minecraft budget toward moderators and customer help. This is because the primary audience is children and it is an MMO - so Microsoft feels obliged to moderate and keep the content safe (and I don't blame them). So this is a big example of when you do need real humans to moderate content. Although some may think an AI could effectively moderate I don't think we are there yet.

Does this concern me about privacy with the advancements in AI and the possibility of a human listening in? Well, I've assumed that a human is going to see my digital footprint for a long time. I speculate that services that allow a user to share in the profits of how their data is used will outwin the current models - which are all just take take take. These companies rationalized that their "service" is just so great and awesome that we, the users, will freely give over data to use their fantastic service. Thing is, more people are realizing their data is actually more valuable than these crappy companies are trading it for. Companies will need to offer much more in return for the data I think and as a result they will be obliged to protect that data since they "paid" so much for it and will treat it more carefully. Just a thought. Sorry for the ramble.