Meta Expands AI Detection for Child Exploitation Content

Meta is expanding its use of artificial intelligence to detect child sexual exploitation material and ads that direct users toward illegal content outside its platforms.

The move comes shortly after Ofcom launched an investigation into Meta over Instagram Instants safety risks. The company has also faced EU scrutiny over child safety failures on Facebook and Instagram in the past.

According to Meta, its new systems analyze both ads and the websites they lead to. This allows Meta to detect apparently harmless ads that covertly direct users toward illegal material.

Meta is using AI to detect hidden signposting

A new LLM-based system detects “signposting,” where ads use indirect language or imagery to point users toward child exploitation material outside Meta’s platforms. Meta can then block violating destinations and take action against associated accounts.

The company is also running additional AI-powered sweeps of advertising content and using an AI red-teaming agent to probe its defenses for emerging abuse tactics. Stronger recidivism detection aims to catch offenders who create new accounts after previous bans.

Meta continues using existing detection tools

Meta combines behavioral signals, classifiers, AI models, heuristics, PhotoDNA, and other image and video matching technologies to detect harmful material. It also blocks links to violating websites and removes posts, comments, and ads containing them.

Between January and June 2026, Meta says it took action against 33.2 million pieces of child exploitation content across Facebook and Instagram. More than 97% was detected before users reported it.

Meta is sharing cases with law enforcement

Meta says it disables violating accounts and reports relevant material to the National Center for Missing and Exploited Children and law enforcement. Since September, it has also reported child-safety cases directly to India’s National Cyber Crime Reporting Portal.

Child safety regulation remains under pressure in Europe as well. A proposed EU KIDS Act could ban social media accounts for children under 13.