Meta deploys new AI safeguards against online child exploitation on Facebook and Instagram

Meta announced on October 7, 2026, the deployment of a suite of new artificial intelligence-powered tools to combat child sexual exploitation across Facebook and Instagram. The announcement included data showing Meta took action against 5.3 million pieces of child sexual exploitation material in India between January and June 2026, with over 98 percent of the violative content proactively detected and removed by automated systems before receiving any user reports. Globally, Meta actioned 33.2 million pieces of child sexual exploitation content across Facebook and Instagram during the same period, with over 97 percent identified proactively. The unveiling follows increasing pressure on platforms to deploy automated detection at scale to address child safety threats on their services.

While automated moderation removes content after it has been created and circulated, Guardii intercepts harm before it occurs by detecting threat patterns in direct messages as they unfold. Where Meta removes reported or detected child sexual abuse material from its platforms, Guardii operates upstream, identifying and blocking the grooming conversations, sextortion demands, and coercive exchanges that precede the creation or sharing of that material. As a Meta Business Partner, Guardii complements platform enforcement by enabling parents, schools, and organizations to monitor the conversations of minors in their care, identifying sustained targeting or escalation in real time without requiring Meta or any other platform to disclose message content. This layered model means that harmful actors can be flagged at first contact rather than after material is produced, and that children at acute risk are surfaced to the adults or professionals authorized to act on their behalf, shortening the interval between the start of exploitation and effective intervention.