`
Why this is noise: A compliance-driven product update triggered by legal pressure, not a capability shift. Nothing here changes how Meta's platforms work for the vast majority of users today.
`
Key Takeaways
- Meta is using AI to scan photos and videos for visual cues including height and bone structure to identify users under 13 on Facebook and Instagram
- The system is not facial recognition, Meta says, and does not identify specific individuals
- Teen Accounts are expanding to 27 EU countries and Brazil on Instagram, and to Facebook in the U.S. for the first time
- The announcement follows a $375 million jury verdict against Meta in New Mexico for misleading consumers about child safety on its platforms
- Meta has threatened to shut down services in New Mexico rather than comply with proposed platform changes it calls “impossible”
Meta announced Tuesday it will use AI to scan photos and videos for visual cues about a user’s age, including height and bone structure, as part of expanded efforts to identify and remove accounts belonging to children under 13 on Facebook and Instagram.
“This is not facial recognition,” Meta said in its announcement. “Our AI looks at general themes and visual cues, for example height or bone structure, to estimate someone’s general age; it does not identify the specific person in the image.”
The visual analysis system operates alongside existing tools that scan text across posts, comments, bios, and captions for contextual clues such as birthday celebrations or mentions of school grades.
Accounts flagged as potentially underage are deactivated and users must verify their age to avoid permanent deletion. The visual analysis feature is currently live in select countries, with a broader rollout planned.
Posts from the community on Reddit
Meta executives knew their products harmed children, disregarded warnings from their own employees, and lied to the public about what they knew. — Raúl Torrez, Attorney General, New Mexico
The announcement arrives weeks after a New Mexico jury ordered Meta to pay $375 million in civil penalties for misleading consumers about platform safety and endangering children.
It is the first jury verdict of its kind against Meta over harm to young people, and a second phase of the case covering public nuisance claims is now underway, with prosecutors seeking court-mandated changes to Meta’s platforms including stricter age verification and default privacy settings for minors.
Meta has since threatened to shut down its services in New Mexico rather than meet what it calls impossible compliance requirements, including a proposed 99% accuracy standard for verifying that child users are at least 13.
A Los Angeles jury also found Meta and YouTube liable for harms to children last month. Meta faces more than 40 state attorney general lawsuits over claims it contributes to a mental health crisis among young people. The New Mexico case is the first to reach trial, and its outcome is being watched closely by every other state pursuing similar claims.
On the product side, Meta is expanding its Teen Accounts system to 27 EU countries and Brazil on Instagram, and to Facebook in the U.S. for the first time, followed by the UK and EU in June. Teen Accounts restrict DMs to existing connections, hide harmful comments, and set profiles to private by default. Meta says hundreds of millions of teens have already been enrolled since the system launched in 2024.
Meta is also calling on legislators to require age verification at the app store and operating system level rather than at the individual app level, a position backed by 88% of U.S. parents according to the company. The argument is that centralised verification would be more consistent and privacy-preserving than requiring every app to build its own system.
Whether Tuesday’s announcements are enough to satisfy courts, regulators, and the 40-plus state attorneys general pursuing Meta will become clear as the New Mexico bench trial proceeds and other cases advance through the courts, a regulatory shift tracked on Relve, an AI trends intelligence platform.
