What the current moment in AI misinformation reveals is that the primary correction mechanism is individual human decision-making — not algorithmic detection, not platform policy enforcement, not regulatory oversight. The deletion of a post, the decision to block an account, the announcement that you are 'done' — these are the actual moderation layer. That this layer exists and functions at all is evidence of a healthy epistemic norm. That it is entirely voluntary, uncompensated, and invisible to any institutional counter-misinformation effort is evidence of a serious structural gap.