Meta Deploys AI to Detect Ads Leading to Child Abuse Material
The company says it acted on 5.3 million pieces of child sexual exploitation content in India during the first half of 2026.
Topics
News
- Fired OpenAI Researchers Warn Dismissals Could Chill Safety Work
- ARTEX Developer Closes AI Hacking Tool After Bank Attacks
- AI Redraws Global Race for Quant Talent as Coding Tools Lower Entry Barriers
- AI Dispatch | India Draws Billions Into AI
- AI Bots Now Make Up Half of Traffic to US Fed’s FRED Database
- Musk Blames ‘Oligarchs’ For Delay in Starlink’s India launch
Meta has introduced new AI systems to detect advertisements that appear harmless but direct users to child sexual abuse material outside Facebook and Instagram, months after such ads drew scrutiny in India.
The company said it acted on 5.3 million pieces of child sexual exploitation content on Facebook and Instagram in India between January and June 2026, with more than 98% detected before users reported it. Globally, it acted on 33.2 million pieces during the period, with more than 97% found proactively.
The new systems target what Meta calls “signposting”, in which an advertisement contains no obvious prohibited material but directs users to illegal content elsewhere.
“An ad can look harmless on its own while directing people towards illegal material hosted off our platforms in a covert way to avoid detection,” Meta said in an October 7 company post.
Meta said it has introduced a large language model to detect such “signposting” and improved its systems for assessing where an advertisement leads rather than examining only its visible content. AI-driven sweeps are also being used to identify material missed by earlier detection systems.
The company has deployed a red-teaming AI agent to probe its own defenses for weaknesses and said it has strengthened systems designed to identify previously banned users attempting to return through new accounts.
Meta continues to use PhotoDNA and other image and video-matching technology alongside behavioral signals and automated rules. It said identified hashes are shared with participating companies through the Tech Coalition’s Lantern program, while links to external sites hosting prohibited material are blocked.
The measures follow regulatory scrutiny in India. In July, the government said it had taken note of reports alleging advertisements linked to child sexual abuse material on social media and had sought a detailed report from the intermediary concerned.
The National Commission for Protection of Child Rights also issued a notice to the platform, according to an official government statement.
Meta acknowledged in July that Instagram ads in India had violated its child exploitation policies and said it had removed additional ads, disabled accounts and blocked related URLs following an investigation.
The company said it reports confirmed material to law enforcement through the US-based National Center for Missing & Exploited Children. It also announced in September that child-safety cases would be reported directly to India’s National Cyber Crime Reporting Portal run by the Indian Cybercrime Coordination Centre.
