Meta adds AI checks for child abuse ads, actions 5.3 million items
Guardii Analysis
Meta deployed AI detection systems to identify child exploitation advertising on its platforms in India and actioned 5.3 million items, according to reporting this week. The move follows scrutiny over whether the company reported CSAM found in India directly to Indian police as required under the Protection of Children from Sexual Offences Act before the September 2026 reporting arrangement was established.
Advertising moderation addresses supply-side signaling, but the demand conversation happens in private messages. Detection built for direct messaging from Guardii identifies the initiating contact from an adult soliciting a minor, the language escalation consistent with grooming or sextortion, and attempts to move a conversation off-platform to unmonitored channels where abuse material is requested or traded. Backed by Startmate and operating as a world-leading online safety platform, it flags the pattern and seriousness of hostile contact in real time across Instagram, Snapchat, Discord, Roblox and other services, surfacing a child in crisis to the adult responsible before an exchange progresses from advertisement exposure to direct exploitation.