Attacks on trust have been around for as long as we have had trust. Generative AI makes some attack vectors easier but it's nothing new: if your trust is earned using a voice you recognise then you model for trust has been broken since before most of us were born.