The convergence across these studies is a serious and urgent finding. People trust AI too much, but more than that, it’s clear that AI fluency creates a new failure mode: Wrong answers delivered in flawless prose get accepted. And the more you are predisposed to trust AI, the worse the problem gets.