The Importance of Trust in Conversational AI

I think with conversational AI, too, you have this problem of fraud happening without someone trying to defraud someone. There's just this innate quality to conversation that we tend to anthropomorphize it. We want to connect with it because of talking to us. I think things like I understand, though, that's clearly deceptive. It probably has this obligation to wake up that person and say, what was the comment?