Global AI Age Detection
Meta is rolling out its artificial intelligence age-estimation technology worldwide on Instagram to identify teenagers who entered false birth dates, notebookcheck.biz reported.
The system automatically shifts suspected underage users into protected teen accounts. Millions of young users will gain these safeguards, according to Meta’s November 9 announcement. Meta announced on November 9, 2026, that it is expanding its AI age-estimation technology worldwide, a step intended to shift millions of additional teenagers on Instagram under the protection of teen accounts. Meta has described this approach in its announcement: Nous utilisons la technologie de l’IA pour identifier les personnes que nous soupçonnons d’être des adolescents, même si elles utilisent une date de naissance d’adulte, et les plaçons sous la protection des comptes adolescents.
The system identifies accounts that appear to belong to individuals aged 13 to 17, even when the account holder has entered an adult birth date, and then limits who can contact them, the content they see, and the time they spend in the application. These protections no longer depend exclusively on the age entered at registration, meaning an account declared as belonging to an adult can still be classified as belonging to an adolescent, though Meta has not provided any overall accuracy figures or false positive rates for the expanded system.
The first wave of protections dates back to 2025, when age verification updates were launched in the United States, the European Union, and Brazil. Meta extended the tool to the 27 EU countries in May, without specific details for France. In August 2026, Meta reached agreements with nearly all US states and territories that had filed lawsuits regarding the effects of its platforms on adolescents, committing to better identify teenagers and children under 13.
Reports diverge on the amount of the August settlement: The Verge suggests 17.1 billion dollars, while Mediafax, relaying Reuters, mentions up to 18 billion dollars. Both versions indicate that Meta agreed to strengthen its age verification practices and impose additional measures.
In October 2026, Meta also concluded a settlement with an adolescent plaintiff in a Los Angeles lawsuit who alleged that Instagram and Facebook were designed to be addictive and caused mental health problems and an eating disorder. The service is free and nothing needs to be installed in France.
How Meta’s AI Scans Profiles
The AI examines profile information, posts, comments, and captions for references to school years or birthday parties. Visual analysis of photos and videos provides extra clues, notebookcheck.biz reported.
Meta states this process does not use facial recognition and does not create biometric identity profiles. Accounts identified as belonging to users aged 13 to 17 switch to teen settings, while accounts suspected to belong to children under 13 face deactivation until users prove their age, gagadget.com reported. The tools examine other profiles and the types of content an account interacts with, birthday messages it receives, and general themes and visual clues pulled from published photos and videos, including height and bone structure.
An early version of an Engadget article also mentioned height or bone structure in photos, but the outlet retracted this, noting those signals are not part of the method described by Meta.
Accounts deemed to belong to children under 13, which is the minimum age for Meta’s social accounts, are removed from the platform. Meta does not provide reliability rates or the number of accounts already switched, and CyberInsider notes the absence of explanations regarding the appeals procedure and data retention duration.
Meta plans to expand protections across platforms
Meta plans to expand similar protections across Facebook and introduce the system on Threads later in 2026, notebookcheck.biz reported. Parents in the United States began receiving Instagram information in November explaining the company’s age-determination approach.
In Australia, Meta started using Apple’s age-range signal as an additional assessment method during device setup, notebookcheck.biz reported. Parents in the United States will see an Instagram video in their feed explaining how age-appropriate protections work and how they can help confirm their child’s age, with 9to5Mac and Mediafax placing the initial parent notifications in November 2026, while The Verge places them in December 2026.
Automatic age controls reached US Facebook accounts in May, and Meta has not specified whether they will expand to other regions. Users aged 16 or younger cannot modify these settings without permission from a parent or guardian, and parents can supervise accounts to set time limits and monitor usage.
In Australia, when a parent configures a child’s phone, the app store can confirm the user’s age and assign an age bracket that participating app developers can access. Meta uses this signal as an extra layer alongside its own AI system rather than replacing the existing system, though the company has not announced the same approach for Europe.
What restrictions apply to Instagram Teen Accounts
What exact restrictions apply to Instagram Teen Accounts?
Teen accounts are private by default and restrict direct messages so that users can only receive messages from people they already follow or connect with, gagadget.com reported. Harmful comments are also automatically hidden. Teen accounts also limit the people who can contact young users, the content they can access, and certain aspects of their activity on Meta’s platforms.
What specific data does Meta’s AI examine on profiles?
The system evaluates profile details, posts, comments, and captions containing references to school years or birthday celebrations, alongside visual cues from published photos and videos, notebookcheck.biz reported.
How does the Apple age signal work in Australia?
When parents configure a smartphone for an adolescent, they can assign an age bracket that participating app developers can access, notebookcheck.biz reported. Meta uses this signal as an extra layer alongside its own AI system.
Keep reading