Red-teaming AI agent, LLM detection and more: Meta shares additional measures to fight child exploitation across Facebook and Instagram

Meta has unveiled a slew of enhanced protective measures against child sexual exploitation on Facebook and Instagram, including autonomous red-teaming AI agents and Large Language Model (LLM) detection tools. The goal is to break up sophisticated predator networks and target elusive advertising schemes that try to direct users to illegal material hosted outside of its platforms. Meta said it took action on 33.2 million pieces of child sexual exploitation (CSE) material globally between January and June 2026, with more than 97% proactively detected by the company before being reported by users.