Instagram Now Guesses Age Worldwide: What a Teen Flag Changes, What Is Known About Undoing It

In 2022, Meta added a check to Instagram. According to its post that year, anyone who tried to change their date of birth from under 18 to 18 or over had to verify their age using one of three options. The birthday on the account was the starting point, and moving it up took proof. Meta's November 2026 announcement no longer takes that birthday as the starting point. Meta says it will use AI to "find people we suspect to be teens, even if they use an adult birthday," and place them in Teen Account protections, the restricted setup Instagram uses for teenagers. The technology has already been running in the US, the European Union and Brazil, according to Engadget. Instagram is now taking it worldwide, the Facebook rollout continues, and Meta says that "later this year, we'll begin launching similar technology globally on Threads."

Two terms carry the rest of this piece, so I will define them once. Age estimation is a guess, in the same sense that a camera's auto exposure is a guess: software reads what is in front of it and decides what is probably right. Age verification is proof, such as an ID document. The new system does estimation. The only route back to an adult account in my sources is the 2022 verification process, and my sources contain no appeal route written for accounts the AI has flagged. Meta's own help pages are the place to look for one. They are not among my sources, and I have not checked them for this piece.

What the system reads, and what a flag switches on

Engadget lists bone structure analysis alongside mentions of school grades or birthday celebrations. TechCrunch's report sorts the signals into two groups: contextual clues, meaning birthdays and school grades found across posts, comments, bios and captions, and visual cues such as height or bone structure. In Australia, Meta has also started using Apple's age-range system as an extra signal, and only Australia is named in my sources for that. No error rate has been published in any of my sources, including a false-positive rate, which is how often an adult is wrongly tagged as a teen. Nothing in them says how the signals are weighted against each other, or whether one birthday caption is enough or the system needs a pattern across many posts.

Meta's announcement says Teen Accounts automatically limit "who can contact them, the content they see, and their time on our apps," and the official post says nothing more about what changes. Engadget is more specific and lists four restrictions: messages from strangers are blocked, the account is private by default, time limits apply, and parental controls apply. Put together, the two sources cover five areas.

Neither source says what the time limits are or what the content limit covers in practice, and the content line is the vaguest thing in the announcement. As for who ends up behind these limits, the group Meta names is teenagers who entered an adult birthday. Meta also says it is investing in technology to "proactively catch accounts that may belong to children under 13," work it ties to its agreement with 52 bipartisan attorneys general. The other group is adults the system misreads, and here my sources are thin. They do not say how many adults Instagram has flagged, or how a flag affects an adult account beyond the restrictions in the table. The evidence that systems like this misread adults at all comes from YouTube, covered below. Engadget also reports that parents in the US get a notification on Instagram explaining how teens are placed into these accounts.

Proving an age, and how YouTube handles the same problem

The 2022 post names the three options: an ID, a video selfie assessed by Yoti, or social vouching, which means other people confirming your age. The November 2026 post, as it appears in my sources, describes no separate route for accounts the AI has flagged, so I cannot confirm that a flagged adult will see these three. What the 2022 post documents is the move from under 18 to adult on Instagram, which is the change a misread adult needs. The options differ in what they ask of you, and none of my sources ranks them. An ID is a document, so it is proof, and it shows your name, your photo and your exact date of birth. The Yoti route is a different kind of thing. Yoti is the company Instagram uses for facial age estimation, which estimates a person's age from a selfie. It is still estimation, so for an adult already flagged by one estimate, this route asks a second estimate to overrule the first, and my sources hold no accuracy data for Yoti's estimate near the 18 line or anywhere else. I have not confirmed what happens to the selfie after the estimate. Social vouching depends on other people, and my sources do not explain how Instagram runs it today.

Instagram is not alone in this. YouTube uses machine learning age estimation in the US, meaning software trained on patterns in data decides how old an account holder probably is, whatever birthdate was entered. Android Central's headline on the subject reads "YouTube's age verification system is back to calling adults kids." Adults who are wrongly flagged there have to verify again with an ID, a credit card or a selfie.

YouTube's list for wrongly flagged adults includes a credit card, and Instagram's 2022 list does not. That comparison sets a current YouTube process against an Instagram process from 2022 that was written for birthday changes, not AI flags, so it shows what each company has documented rather than what a flagged Instagram user will be offered today. YouTube's record is the clearest public evidence in my sources that these systems mislabel adults. I have no comparable figure for Instagram, in either direction.

What I would check before handing over proof

• Whether you have been told anything. My sources do not say whether Instagram notifies a flagged user. The only notification they mention is the one Engadget says US parents receive.
• Whether Meta's current help pages describe a route written for AI flags, since the announcements in my sources do not.
• What happens to the selfie after the estimate, which my sources do not confirm, before choosing the Yoti option.