Meta announced on Monday that it will expand its AI-powered age detection technology across Facebook and Threads. The system uses visual and contextual analysis to identify users who may be teenagers, even when they register with an adult birthday. Until now, the technology operated primarily on Instagram. This expansion builds on Meta’s agreement with 52 bipartisan US attorneys general to strengthen how it identifies teens and children under 13.
The results so far suggest the system works at scale. According to Meta, it has enrolled at least 54 million teens into Teen Accounts globally. Among teens aged 13 to 15, 97% chose to remain in those protections after enrollment. The company also recently expanded Teen Account protections to Facebook and Messenger.
Meta talks about this new technique as:
We use AI technology to find people we suspect to be teens, even if they use an adult birthday, and place them in Teen Account protections. Since first introducing this technology, we’ve been placing more teens in age-appropriate protections and are encouraged by the results.
However, questions about actual effectiveness remain. In Australia, Meta removed more than 750,000 profiles it detected as belonging to underage users. That sounds significant. On the other hand, a report from Australia’s eSafety Commission published in August found that 81.5% of underage users still accessed restricted apps despite the bans. That gap suggests either the detection systems miss many users or teens find workarounds faster than platforms can block them.
The AI system analyzes photos and videos for visual cues like height, bone structure, and facial proportions. According to multiple reports, it also scans contextual signals such as birthday posts and interaction patterns. Meta has emphasized that the technology does not use facial recognition. Instead, it estimates age through structural and thematic analysis. If the system determines an account may belong to someone underage, it places the account into Teen Account protections with restricted content, limited messaging, and default privacy settings.
For Australia specifically, Meta announced it will now incorporate Apple’s age-range signal into its detection process. When a parent sets up a smartphone for their teen, the app store can confirm the user’s age range. Apple then passes that information to developers like Meta each time the teen downloads an app. Meta called this a voluntary adoption and again urged app stores to take a larger role in age verification at the download level.
The expansion to Threads represents a new front. Meta said it will begin launching similar detection technology on Threads globally later this year. This matters because Threads has grown rapidly and currently lacks the same level of teen-specific protections that Instagram has built over the past two years.
Meta also used Monday’s announcement to call out competitors directly. The company urged TikTok and YouTube to adopt the same practices for teen safety. This framing drew criticism from observers who noted that Meta would likely not have implemented many of these protections without the attorneys general lawsuit. In August, Meta agreed to pay $18 billion and adopt new teen safety measures to settle that landmark case. The settlement included time limits on teen usage, notification blocks during school hours and overnight, and the option for teens to set non-algorithmic feeds as default.
On a blog explaining the new development, Meta used this feature video loop to tell the world of the endgame of this plan:
The broader regulatory picture continues tightening worldwide. Meta, TikTok, and Google recently pushed back against a California law targeting platforms over teen safety. Meanwhile, the EU Digital Services Act imposes its own scrutiny on age verification practices. Australia’s under-16 social media ban went into effect in December 2025, and other countries continue weighing similar restrictions.
For Meta, the calculation remains straightforward. Expanding AI detection costs money and creates friction. However, not expanding it risks more lawsuits, larger fines, and stricter government-imposed rules. The $18 billion settlement proved that the cost of inaction far exceeds the cost of building better safeguards.
