Source: Thesignal
Substack's move to add visible AI-detection tooling signals that platforms can no longer ignore reader anxiety about authenticity without appearing complicit—but the feature itself is a half-measure that likely catches obvious slop while missing sophisticated synthetic content. The friction point is whether readers trust that human judgment (editorial standards, author reputation, community norms) still carries weight. Substack's technical band-aid doesn't restore that trust. Content moderation shifted from "Is this allowed?" to "How do we rebuild trust after the tools failed?" Here, the tool being surfaced is detection itself, making the problem visible in a way that may train readers toward skepticism rather than reassurance.