Meta puts sneaky teens into safer accounts with visual AI

Lying about your birth year to bypass internet restrictions used to be a rite of passage. If you grew up online, you probably punched in 1990 just to get past a sign-up gate.

Meta is making sure today's kids cannot pull that off quite so easily.

Meta announced a major expansion of its AI-powered age assurance tools. The goal is simple: catch teenagers pretending to be adults and proactively funnel them into restrictive Teen Accounts.

The feature has already been live for Instagram users in the US, the UK, Canada, and Australia, placing millions of accounts under tighter settings. Now, Meta is bringing it to 27 EU nations and Brazil. Facebook is also getting the tool for the first time in the US, with rollout to the UK and the EU scheduled for June.

Context clues and controversial visual scans

To sniff out underage users, Meta's AI does not just check what date is saved in a profile. It scans entire accounts for contextual evidence.

The system reads captions, comments, and bios looking for telltale signals like birthday party milestones or references to school grades. Meta is even expanding this monitoring into Instagram Reels, Instagram Live, and Facebook Groups.

More notably, Meta is rolling out visual analysis in select countries to scan photos and videos for age cues when written indicators are missing.

That part quickly triggered privacy alarms. Meta added a post-publication correction on July 31 specifically clarifying that the system is not facial recognition. According to the company, it does not build facial fingerprints or identify individuals. Instead, it looks for general visual patterns that suggest a minor might be in the frame.

If the AI flags someone as under 13, the app deactivates the account until the user submits proof of age. Meanwhile, anyone attempting to change their birthday from under 18 to adult status must verify themselves with a government ID or third-party selfie age estimation from Yoti.

Meta is also sending alerts to US parents this month with tips on discussing age honesty, while using AI models alongside human reviewers to process community reports faster.

Shifting the burden to Apple and Google

Meta is not shy about pointing fingers at where it thinks this responsibility truly belongs.

In the announcement, Meta pushed for legislation requiring app stores to verify user age at the operating system level and pass those signals down to individual apps. To back up its argument, Meta cited internal data showing that 88% of US parents support putting verification onto the app stores instead of individual platforms.

Why it matters

For teens and parents, this crackdown will make dodging safety walls noticeably harder. Teen Accounts automatically limit who can message minors and restrict the types of sensitive content they can encounter. The real headache will be false positives: young adults who get mistaken for minors will have to go through the friction of uploading IDs to regain normal access.

For the broader tech ecosystem, Meta's lobbying push is a clever political play. Regulators worldwide are coming down hard on social platforms over child safety. By advocating for app store-level verification, Meta is trying to offload both the blame and the compliance burden onto Apple and Google.

We think Meta's multimodal approach makes tactical sense because simple text fields have always been easy to spoof. However, analyzing visual cues in personal photos to guess someone's age sits uncomfortably close to biometric profiling, no matter how Meta frames its disclaimers. In strict privacy regions like Europe, this technology is almost certain to attract regulatory friction.