Meta Platforms will expand its AI-powered age-estimation technology globally, targeting millions of additional teenagers on Instagram who lie about their dates of birth. The company announced the move in a statement on Monday, framing it as a decisive step in online safety rather than a concession to mounting regulatory pressure.

The system identifies users suspected of being between 13 and 17 years old even when their account information lists them as adults. Once flagged, these users are automatically enrolled in age-appropriate account protections, removing the choice from the teenager entirely.

What does the technology actually examine? According to Meta, the AI analyses a broad set of signals: profiles, posts, comments, biographies and captions, including contextual clues such as references to school grades and birthday celebrations. It also uses visual indicators in photographs and videos to estimate a person's general age.

The company was quick to close one obvious line of attack. Meta stated the process does not involve facial recognition or identifying individuals in images. That distinction matters — facial recognition would trigger far stricter legal scrutiny in most jurisdictions, and Meta's wording leaves open how accurate visual age estimation is without it.

What protections do flagged teens receive?

The Teen Accounts feature, available on Instagram, Facebook and Messenger, imposes a set of default restrictions:

• Private account settings by default
• Restrictions on direct messages from people the teenager does not follow or already know
• Limitations on sensitive content
• Reminders encouraging users to take breaks from the platforms

Meta said the global expansion follows encouraging results from an earlier deployment, which placed more teenagers in age-appropriate digital environments despite attempts to circumvent age restrictions. The company did not publish specific figures on how many misaged accounts the earlier system caught — "encouraging results" is a claim without a number attached, and the burden of proof remains Meta's.

Why is Meta doing this now?

Determining users' actual ages remains a major challenge for social media platforms, as some teenagers deliberately provide older dates of birth to access adult features. Meta has been investing in AI systems that examine account information and activity for indicators suggesting a user is underage.

The company said the expansion builds on its recent agreement with 52 bipartisan attorneys general across US states, territories and the District of Columbia to strengthen identification of teenagers and children under 13 on its platforms. It also arrives amid growing international scrutiny of social media companies over children's safety, harmful content and the effects of excessive use on young people's wellbeing — and amid ongoing legal pressure on Meta over allegations concerning the impact of its platforms on young users. Meta's statement did not address that litigation directly.

Where and when does it roll out?

The rollout begins on Instagram, covering millions of additional teenagers. Meta said it is continuing to extend the technology to Facebook, while a similar global rollout on Threads, its text-based social networking platform, is expected later this year.

Two regional additions stand out:

• United States: parents will begin receiving Instagram notifications explaining how Meta determines whether teenagers are placed in age-appropriate accounts, and how families can help ensure ages are correctly represented.
• Australia: Meta has started using Apple's age-range signal as an additional verification method. The system lets app developers receive an age range from Apple when users or their parents opt to share the information through the device's app store.

The Apple route is notable because it relies on declared, opt-in data rather than behavioural inference — a cleaner privacy posture than profile and photo analysis.

What does this change for privacy?

The expansion underscores the broader dilemma facing technology companies and regulators: effective age assurance (the practice of verifying a user's age, here through inference rather than documents) increasingly depends on personal information, online activities and visual content. Every protection Meta promises here rests on deeper surveillance of its users' behaviour; the company offers no detail on retention limits or error rates for its estimates.

Meta said it will continue working with parents, technology partners and other industry stakeholders to improve age assurance, and will keep investing in technology to identify accounts that may belong to children below its minimum age requirement. What happens when the AI gets a user's age wrong — and whether mistaken adults can appeal an automatic downgrade — is a question the statement does not answer.