Meta said it removed or took action against 5.3 million pieces of child sexual exploitation content across Facebook and Instagram in India between January and June 2026. The company is increasingly using artificial intelligence to identify abusive material as well as links that may direct users to illegal content hosted outside its platforms.
According to Meta, more than 98% of the child sexual exploitation content detected in India during the six-month period was identified and actioned before it was reported by users.
Worldwide, Meta said it took action against 33.2 million pieces of such content during the same period, with more than 97% identified proactively by its systems.
Meta Expands AI-Based Detection Systems
Meta said individuals involved in online abuse are increasingly using advertisements and apparently harmless posts or content to guide users towards child sexual abuse material (CSAM) hosted on external websites.
To address this tactic, the company has introduced new detection systems based on large language models. These systems are designed to identify what Meta describes as “signposting” — content that may appear harmless by itself but could be used to covertly direct people towards illegal material.
The company is also examining the destination of advertisements instead of assessing only the content displayed within an ad. Meta said it has introduced AI-powered sweeps of advertising content along with a red-teaming AI agent designed to identify weaknesses in its detection systems.
The company has further strengthened measures to identify repeat offenders who attempt to return to its platforms by creating new accounts after previous accounts have been removed.
Photo Matching and Behavioural Signals Used
Meta continues to rely on technologies including PhotoDNA and other image- and video-matching systems to identify known child sexual abuse material across Facebook, Instagram and Threads.
Its detection systems also consider behavioural patterns, metadata and automated classifiers when identifying potentially suspicious activity.
When a link leading to an external website containing violating material is identified, Meta can block the link. The company can then search its platforms for related advertisements, posts and comments that contain the same link.
Meta to Report Cases Through India’s Cybercrime Portal
Meta said it will begin reporting child safety cases directly to the National Cyber Crime Reporting Portal operated by the Indian Cybercrime Coordination Centre (I4C).
The move adds to Meta’s existing process for reporting child sexual abuse material to law enforcement agencies, which is carried out with assistance from the National Center for Missing & Exploited Children.
The company said its expanded use of AI, combined with image matching, behavioural analysis and strengthened reporting systems, is aimed at identifying abusive content and related activity before it is reported by users.