Meta on Wednesday said it has deployed new AI tools to strengthen its ad review systems and detect and disrupt child sexual exploitation on Facebook and Instagram, as it works to counter increasingly sophisticated tactics used by online predators.
The company said the new measures are designed to identify advertisements that either feature child sexual abuse material (CSAM) or covertly direct users to violating activity hosted outside its platforms.
Meta acted on 5.3 mn pieces of child abuse content in India
Meta also disclosed that it acted on 5.3 million pieces of child sexual exploitation content in India in the first half of 2026. "Between January and June 2026, Meta actioned 5.3 million pieces of child sexual exploitation content on Facebook and Instagram in India, with more than 98 per cent of the content identified and addressed proactively before it was reported," the company said in a blog post.
Globally, Meta actioned 33.2 million pieces of such content during the same period, with more than 97 per cent detected proactively.
The company said predators are increasingly using advertisements to evade detection, including by creating seemingly harmless ads that direct users to illegal material hosted on external websites. Meta said it has expanded its investigations beyond individual reports to identify related activity across its platforms, disable accounts involved and block associated off-platform links.
As part of the latest measures, Meta has introduced large language model-based detection to identify "signposting" in advertisements, where seemingly benign content is suspected of directing users towards illegal or harmful material.
AI-driven sweeps, red-teaming AI agent
The company has also enhanced its ability to assess where advertisements lead, rather than examining only their visible content. This allows Meta to block violating destinations and take action against the accounts responsible.
Meta said it has introduced AI-driven sweeps of advertisements to identify child sexual exploitation content that may have been missed by earlier systems. It has also deployed a red-teaming AI agent to test its own defences and identify vulnerabilities that could be exploited by predators.
The company has further strengthened its systems to detect repeat offenders who attempt to create new accounts after being removed.
Meta said its broader child-safety systems use behavioural signals, PhotoDNA and other image- and video-matching technologies, automated classifiers and heuristics-based rules to identify emerging patterns of violations.
"Keeping young people safe on our apps is among the most important work we do, and it never stops," Meta said.
The disclosure comes amid a government crackdown on child sexual abuse material.
In July, the IT ministry issued a notice to Meta over such ads on Instagram, ordering them to be disabled, and summoned its global executives to India.
In India, Meta also said it would report child safety cases directly to the National Cyber Crime Reporting Portal operated by the Indian Cybercrime Coordination Centre (I4C). The company said this builds on its existing reporting of CSAM to law enforcement through the portal with assistance from the National Center for Missing & Exploited Children.
(With ANI, PTI inputs)
Also Read
Meta Muse AI Agent arrives on Mac with support for complex tasks: What to know?
