Meta systems fail to detect hundreds of AI-generated child abuse ads

Since early August, Meta's automated moderation systems have failed to detect and prevent the spread of more than 250 advertisements containing child sexual abuse material generated by artificial intelligence. The presence of these ads across Meta's platforms was uncovered by an investigation conducted by the Tech Transparency Project and shared with WIRED.

The findings highlight ongoing vulnerabilities in Meta's content moderation and advertising review processes regarding the proliferation of AI-generated illicit material. Despite existing policies against child exploitation, the system's inability to catch these hundreds of advertisements raises significant concerns about platform safety and the enforcement of digital content standards.