Meta reported acting on 33.2 million pieces of child sexual exploitation content on Facebook and Instagram in the first half of 2026, with more than 97% detected by its own systems before users flagged it. In India, the company acted on 5.3 million pieces during the same period, with over 98% found proactively.
The company is now rolling out AI tools aimed at a newer tactic: ads that appear harmless but secretly direct users to illegal material outside Meta's platforms. A new large language model system is designed to detect this 'signposting' behavior, and Meta says it now looks at where an ad sends users, not just what the ad contains. Additional AI-driven scans and a 'red-teaming AI agent' are meant to find gaps in Meta's defenses before abusers exploit them.
The announcement comes amid continued scrutiny over child safety. Meta has faced lawsuits and criticism from lawmakers, and in August reached an agreement to pay up to $18 billion to settle a child safety lawsuit involving 29 U.S. states. The company has also introduced other child-safety features this year, including parental controls for Meta AI and preteen accounts on WhatsApp.