Meta Approved Hundreds of Ads Containing AI-Generated CSAM Over Nine Months
Engadget·medium signal
The Tech Transparency Project found hundreds of ads containing AI-generated child sexual abuse material that passed Meta's ad review and ran between November 2025 and August 2026. The failure is in automated ad moderation specifically, a system Meta has been scaling with AI while cutting human review. It is a direct counterexample to the claim that AI classifiers close the moderation gap.