Meta says its upgraded AI systems now scan where advertisements lead users, not just what they show, as it reports acting on 5.3 million pieces of child sexual exploitation content in India in the first half of 2026.
Â
Meta has rolled out new artificial intelligence tools to detect child sexual exploitation content more effectively on Facebook and Instagram, including cases where offenders use innocuous-looking advertisements to steer people towards illegal material hosted outside its platforms.
The company said it took action on 5.3 million pieces of such content in India between January and June 2026, with over 98% detected and dealt with before any user reported it. Globally, Meta acted on 33.2 million pieces during the same period, of which more than 97% was identified proactively.
How offenders exploit ads
Meta said it has uncovered advertisements that look harmless on their own but are built to lead users to illegal content on external websites. Because the ad itself may contain nothing clearly in violation, such cases are hard for automated systems to flag.
To counter this, the company has introduced tools that examine the destination an ad points to, not only its visible content. One system uses large language models to spot what Meta calls “signposting”, meaning ordinary-seeming ad material that could be used to covertly direct people to child sexual exploitation content or other harmful activity.
Meta is also running additional AI-driven sweeps of ads to catch material that earlier systems may have missed. It has deployed an AI “red-teaming” agent that probes its own safety systems to expose weaknesses and uncover new offender tactics. Detection of repeat offenders who open fresh accounts after being removed has been strengthened as well.
What happens after detection
When Meta identifies or is alerted to potentially exploitative content, it investigates. Material that breaches its policies is removed, and accounts tied to malicious activity can be disabled.
The company also uses PhotoDNA and other image and video matching technologies to identify known child sexual abuse material, tools it says it has applied across its platforms since 2011. Links to external sites hosting violating content are blocked, after which Meta searches for and removes other posts, comments and ads carrying the same link.
Identified material is reported to law enforcement through the National Center for Missing & Exploited Children (NCMEC), in line with applicable law. In India, Meta said it will additionally report child safety cases directly to the National Cyber Crime Reporting Portal run by the Indian Cybercrime Coordination Centre (I4C).
Meta acknowledged that tackling child exploitation demands constant updates to its systems, as offenders keep changing tactics and moving across platforms.
