Meta Adds AI Tools to Detect Ads Linked to Child Abuse Material

Meta says it took action against 33.2 million pieces of child sexual exploitation content on Facebook and Instagram during the first half of 2026, while rolling out new AI systems designed to detect ads and accounts that may direct users to illegal material elsewhere online.

According to Meta’s latest child-safety update, it detected more than 97% of the content it acted on globally before users reported it. In India, Meta said it removed or acted on 5.3 million pieces of such content, with more than 98% identified proactively.

Meta is targeting ads that act as gateways

One of the new systems uses a large language model to detect what Meta calls “signposting.” These are ads that may appear harmless on their own but are suspected of directing users to websites or services hosting child sexual abuse material or other illegal content.

Meta says it now analyses not only an ad’s content but also the destination it sends users to. That allows the company to block prohibited destinations and take enforcement action against the accounts responsible for promoting them.

The company is also using additional AI-based scans to identify exploitation content that earlier detection systems may have missed. Meta says it plans to keep adding new signals as it learns more about how abusive networks change their tactics.

AI is also being used to test Meta’s own defence

Meta has introduced a red-teaming AI agent that probes its safety systems for weaknesses that bad actors could exploit. The goal is to identify possible bypass techniques before they become widely used.

The company is also improving systems designed to detect people who return with new accounts after previous accounts have been removed. That work is intended to make it harder for repeat offenders to evade enforcement by simply creating new profiles.

Child safety remains under regulatory scrutiny

The new measures arrive as Meta continues to face legal and political pressure over child safety on its platforms. The company has introduced several additional protections this year, including parental controls for Meta AI, preteen accounts on WhatsApp and alerts for parents when children search Instagram for self-harm content.

WhatsApp also added more parental controls in September, including options to limit teenagers’ use of Channels, restrict who can view status updates, control who can add them to groups and receive notifications about certain group activity.

Meta’s latest approach shows a shift toward using AI not only to find prohibited material directly, but also to identify the surrounding behaviour, links and account patterns that may be used to distribute it.