Meta Extends AI Age Estimation Across Instagram to Move Undeclared Teens Into Protected Accounts
Meta is expanding its AI-powered age-estimation technology globally to find teenagers on Instagram who have not correctly disclosed their age. The rollout will automatically place millions more teens into Teen Account protections, according to The Verge.
The system is not new. Meta launched the AI-powered Instagram age-assurance updates in the US, EU, and Brazil last year. In May, the company said it was extending automatic placement of suspected teens into Teen Account protections in the EU, and Reuters reported that the safeguards would cover 27 EU countries. In the same period, automatic age-assurance checks reached US Facebook accounts. Meta has also begun using Apple's age-range signal as an additional detection method in Australia.
The detection layer infers age rather than verifying it. Instagram's tools look for signals that a user is likely under 18, such as the profiles and content a user interacts with, birthday messages the user receives, and general themes and visual cues from uploaded photos and videos, including height and bone structure. The system acts on those inferences at scale, not on declared birthdates.
Enforcement works in three tiers. Accounts believed to belong to users under 13, the minimum age for Meta social accounts, are removed. Users believed to be 13 to 17 are moved automatically into Teen Accounts, a bundle of stricter settings, even if they signed up with a fake adult birthday. Those accounts are private by default, limit the kinds of content teens can see, and block direct messages from strangers. Teens aged 16 or younger cannot change those settings without a parent's or guardian's permission.
Parental involvement is the other half of the design. Meta's own announcement describes parent-facing controls. Meta's earlier article on enrolling suspected teens, last updated in September 2025, described the same approach of using AI to place teens into Teen Account settings with parental input. Parents in the US will start receiving a notification on Instagram next month, in December, explaining the age-estimation process and how they can make sure their teen's age is accurately reflected.
Meta has also been widening the content side. In June, the company announced new 13+ content settings for Teen Accounts, expanding globally across Instagram, Facebook, and Messenger. Meta says the setting draws on movie rating criteria and parent feedback, and is intended to show age-appropriate content to teens by default.
In my view, the central technical question is classification under uncertainty. Inferring age from behavioral and visual signals is a probabilistic judgment, and the two possible errors land on different people. An adult misclassified as a teen gets restrictions they did not need, while a minor who is missed stays outside the protections. Anyone who has tuned a classifier will recognize the precision-recall tradeoff underneath a global rollout like this one.
The broader context here is that default settings do much of the work. Most teens never open a privacy menu, so a system that starts them in a restricted configuration changes the baseline for everyone it reaches, whatever its error rate. The open questions are how accurately the signals separate age groups across markets with different languages, device mixes, and usage patterns, and how parents respond when the notification arrives. Those answers will show whether automatic placement protects the teens it is meant to protect.
