Home / Crime and Justice / Meta Fights Child Exploitation With New AI
Meta Fights Child Exploitation With New AI
7 Oct
Summary
- Meta removed 33.2 million child exploitation pieces in H1 2026.
- New AI tools detect ads secretly directing to illegal content.
- Over 97% of actioned content was system-detected.
Meta has intensified its fight against child sexual exploitation, announcing that it took action on 33.2 million pieces of harmful content across Facebook and Instagram during the first half of 2026. Crucially, over 97% of this content was identified by Meta's automated systems before any user reports.
The company is now deploying new AI tools to combat a growing tactic where seemingly innocuous ads are used to direct users to illegal child abuse material. A new large language model system, dubbed 'signposting' detection, analyzes where ads lead, not just their content, enabling Meta to block malicious destinations and accounts.
Further enhancing its safety measures, Meta is utilizing AI-driven scans to catch previously missed exploitation content and employing a 'red-teaming AI agent' to proactively test and strengthen its defenses. These enhanced efforts come amid ongoing pressure and scrutiny regarding child safety on social media platforms.
