Demoting borderline content instead of removing it: reach reduction that nobody can audit

Popularity Index: 5 Updated at 2026-10-11 09:25 UTC

Sites

Related Topics

Fact-check labels help some readers and harden others: the backfire debate is not settled Trusted flagger programmes: who gets to tell a platform what counts as false Outside review boards for moderation appeals: oversight that platforms fund themselves AI-generated uploads are flooding recommendation feeds faster than moderation can flag them The first 30 minutes decide whether a new upload gets recommended at all Will the EU's Digital Services Act fines actually change how platforms moderate content? Deepfakes get the headlines but cheap fakes do most of the damage Transparency reports measure the wrong thing: takedown counts say nothing about whether the right posts were removed Section 230 in the US: platforms are liable for what they know about and shielded from everything else Community notes beat top-down labels because they need agreement across viewpoints

Hot Categories

Featured Categories