Meta said on Monday it will stop taking teenagers at their word on Instagram. The company is expanding its Instagram AI age detection system worldwide, a move it says will place millions more teens into Teen Account protections even when they signed up with an adult birthday.
“Today, we’re expanding our AI age-estimation technology globally to place millions more teens in Teen Account protections on Instagram,” Meta said in its announcement. The rollout lands three months after Meta reached an agreement with 52 state attorneys general and settlements worth up to $18 billion over claims its platforms harm young users, according to Reuters.
The birthday box, in other words, no longer decides anything.
Until now, the system ran only in a handful of places. Meta started testing automatic placement in the United States on April 21, 2025, then expanded it to the UK, Canada, and Australia that September. The EU and Brazil followed in May 2026. Monday’s announcement takes it everywhere Instagram operates, and Meta says Facebook comes next, with Threads to follow later this year.
How Instagram’s AI age detection works
Meta’s system doesn’t ask for an ID up front. Instead, it reads a whole profile for clues: birthday messages, mentions of school grades, the accounts a user follows, and the posts they interact with. It also analyzes photos and videos for visual signs that the person in them is young.
Meta is careful about one line. The company says the technology is not facial recognition: it doesn’t create facial fingerprints, and it doesn’t identify who is in a photo. It estimates an age range, and that estimate is probabilistic, which means it can be wrong.
Accounts the system flags as belonging to 13-to-17-year-olds get moved into Teen Account protections automatically. Accounts it suspects belong to children under 13, Instagram’s minimum age, face a harder road: they must prove their age or lose the account.
What Teen Accounts actually do
Teen Accounts are Instagram’s restricted mode for under-18s, and the settings are stricter than most adults realize. Profiles are private by default for teens under 16 and for under-18s when they sign up. Messaging is locked to the strictest setting, sensitive content is filtered, and limits apply to tags, mentions, and offensive terms in comments and message requests.
There’s a time element, too. Teens get a reminder to leave Instagram after 60 minutes each day, and notifications go quiet during sleep mode from 10 p.m. to 7 a.m. Teens under 16 need a parent’s permission to loosen any of it, with supervision tools letting parents approve or deny the request.
Teen Accounts aren’t new. They reached every country where Instagram is available by June 11, 2025. What’s new is the enforcement: instead of waiting for a teen to admit their age, Meta’s AI now hunts for them.
Why Meta is doing this now
The timing isn’t subtle. Meta has spent the year settling and defending a wave of lawsuits alleging its platforms harm teens. Reuters reported Monday that the company faces multiple lawsuits on that front, and that it struck settlements worth up to $18 billion with nearly every US state and territory in August. Last month, Meta settled with a teen plaintiff in a Los Angeles trial that tested claims Instagram and Facebook were designed to be addictive and caused mental health problems and an eating disorder.
The agreement with 52 bipartisan attorneys general, announced August 26 and approved by a judge the next day, requires Meta to invest further in systems that spot under-13 accounts and recognize 13-to-17-year-olds even with an adult birthdate on file. Most of its terms run for ten years, and an independent auditor will test Meta’s compliance annually for five years. The global rollout builds on those commitments, though the agreement’s own protections cover only participating US jurisdictions.
What happens when the AI gets it wrong
Every age-estimation system makes mistakes, and Meta’s announcement doesn’t say how often. The company hasn’t published error rates or appeal volumes, and Monday’s post doesn’t describe a new appeal route. The existing one applies: anyone placed in teen settings who tries to change a birthday from under 18 to 18 or older must verify their age, either with an ID or with a video selfie checked by Yoti, an outside age-estimation firm.
Parents in the United States will start seeing an Instagram in-feed video this month explaining how the placement process works, linked to guidance from Larry Magid, president and CEO of the nonprofit ConnectSafely.
In Australia, Meta has started using Apple’s age-range signal as an extra check: when a parent sets up a phone for a teen, the App Store can pass the device user’s age range to the app. Meta supports legislation requiring app-store age verification, with its own estimation serving as a backstop.
Why this matters
Age assurance on social media has quietly shifted from a form field to an inference problem. For years, the whole system rested on a teenager typing a believable birthday, and everyone involved knew how that went. Meta’s move doesn’t fix the underlying tension, it just moves it: now the question is whether an algorithm can guess a person’s age from their posts and photos, and what happens to the adults it mislabels as kids.
That trade is the real story here. Millions more teens will get protections they would never have asked for, which is the point. But adults wrongly flagged will have to prove their age to a company that already has their data, and Meta hasn’t said how many of them there are. Until it publishes accuracy numbers, the system asks users to trust a guess. If the rollout holds up, regulators will expect TikTok and YouTube to match it. If it misfires, it becomes the next exhibit in the lawsuits Meta just paid billions to settle.
Frequently asked questions
How does Instagram’s AI age detection work?
Meta’s system scans profile signals such as birthday posts and school mentions, plus visual cues in photos and videos, to estimate whether an account belongs to someone aged 13 to 17. Meta says it is not facial recognition and does not identify the person in a photo.
What happens if Instagram thinks I’m a teenager?
The account is placed into Teen Account protections: a private profile by default, the strictest messaging and content controls, and a 60-minute daily time reminder. Teens under 16 need a parent’s permission to loosen those settings.
What if Instagram gets my age wrong?
Meta says users can correct a mistake through the app’s settings. Anyone trying to change a birthday from under 18 to 18 or older must verify their age with an ID or a video selfie checked by the outside firm Yoti.
When is the AI age detection coming to Facebook and Threads?
Meta says it is still rolling the technology out on Facebook and plans to start a global launch on Threads later in 2026. Instagram is the first platform to get the worldwide rollout.
Sources: Reuters, Meta, The Verge