Researchers from Tech Transparency Project identified more than 50 paid advertisements featuring AI-generated child sexual abuse material that ran across Facebook, Threads, Messenger and Instagram between November 2025 and August 2026, according to a report published by Wired. The ads reached thousands of accounts, and several linked to nudify or undressing applications. Meta stated that the majority of the ads had minimal reach and many were disabled before Wired shared them, and that the company recently launched new AI technology to better detect and block violating ads at upload, while also referencing the removal of over 36 million pieces of child sexual exploitation content last year.
Platform-level content moderation, even when augmented by AI scanning at upload, failed to block dozens of CSAM ads that ran for months and reached thousands of users across Meta's properties, illustrating the structural gap between upload filters and in-message interception. Guardii's real-time anti-CSAM detection operates at the transmission layer in direct messages, where the vast majority of grooming, solicitation and material-sharing occurs outside the advertiser and public-post queues that Meta's upload scanners prioritize. For parents, schools and safeguarding organizations, blocking a CSAM image or nudify-app link before it reaches a child's inbox is operationally distinct from flagging it after thousands of impressions; Guardii delivers the former by monitoring the private channels where exploitation material circulates peer-to-peer, rather than relying on platform upload gates that demonstrably let violating content through.
