Meta has detailed a new underage-enforcement system that adds visual analysis to its age-assurance tooling on Instagram and Facebook. In a Tuesday post, the company said its AI now scans photos and videos for broad age cues such as height or bone structure, alongside text and interaction signals, to estimate whether an account likely belongs to someone under 13.

What Meta confirmed

Meta said the system is not facial recognition and does not identify a specific person in an image. Instead, it combines visual cues with signals from posts, comments, bios, captions, and other activity to decide whether an account should be flagged as potentially underage.

If Meta determines that an account may belong to someone under 13, it says the account will be deactivated and the user will need to complete an age-verification process to avoid deletion. The company said visual analysis is available only in select countries for now, while broader AI-based age checks are already in wider use.

Broader rollout

The announcement also expands Meta's teen-protection flow. Meta said its system for placing suspected teens into stricter Teen Accounts is now expanding to 27 EU countries and Brazil on Instagram, and to Facebook in the U.S. first, followed by the U.K. and EU in June.

The conservative takeaway is that Meta is widening automated age estimation across images, video, text, and interactions, but the most aggressive new feature is still in a limited rollout and its real-world accuracy has not been independently validated.