Meta is expanding AI-based age estimation worldwide, starting this month on Instagram, with Facebook and Threads to follow later in the year. The system analyses behaviour and posts to estimate a user's age. It can move suspected teenagers into Teen Account protections even when they entered an adult birthday, which teens often do to get round restrictions. Meta says it has tested the technology in the United States, Britain, Canada and Australia. In the US, parents who set up Instagram for their teens will be notified this month.
Existing Teen Account limits include a ban on children under 13, reminders to leave Instagram after 60 minutes, muted notifications at night, a two-hour daily limit, and the blocking of social comparison features such as seeing likes on other teens' accounts.
The expansion follows legal pressure. In August the company agreed to a settlement with US states under which it will pay $18 billion and improve how it identifies minors. Under-18 users face a midnight-to-6am curfew, push notifications are blocked from 8am to 3pm, cosmetic-procedure filters are restricted by default, and teens receive break prompts after 15 minutes of continuous use. Parents are told when a teen first messages an adult and when a teen searches for suicide, self-harm or eating-disorder topics, and only parents can remove the changes. Attorneys general across the US had sued, alleging that Meta's platforms encourage addiction, fail to verify ages and expose children to harmful content.
Australia passed a ban on social media for under-16s last year, and this month the Digital Freedom Project and Reddit challenged the law in the High Court.
Meta has urged TikTok and YouTube to adopt similar practices.





