Meta today extended the artificial intelligence it uses to spot teenagers posing as adults to Instagram accounts worldwide, placing users it suspects are between 13 and 17 into Teen Accounts even when they signed up with an adult birthday. According to the company's newsroom post, the change will put "millions more teens" under restrictions that limit who can contact them, what they see and how long they spend on the app. Meta gave no figure beyond that phrase.

Two narrower changes came with it. In Australia, Meta has started using Apple's age-range signal, which the App Store passes to developers, as an additional way to determine age. In the United States, parents will begin seeing an in-feed video on Instagram this month explaining how the company places teens in age-appropriate settings. Meta presented the whole package as building on "our commitments under our recent agreement with 52 bipartisan attorneys general to strengthen how we identify teens and children under 13 on our apps."

In Short

Meta now uses AI everywhere Instagram operates to find users who are probably teenagers but registered as adults, and it moves those accounts into Teen Accounts with stricter default settings. A typed-in birthday no longer decides on its own whether an Instagram account is treated as an adult, which changes who can reach those users, what content they get and how their time on the app is capped. Facebook and Threads are next, Australia adds Apple's age signal to the mix, and Meta has still not published how often its age estimates are wrong.

What changes today

According to Meta, Teen Accounts on Instagram, Facebook and Messenger "provide teens with a set of built-in protections that automatically limit who can contact them, the content they see, and their time on our apps." Those protections only reach accounts Meta knows to belong to teens. "Ensuring all teens receive these protections depends on knowing their age," the post states, calling this "an industry-wide challenge as some teens may misrepresent their age online to access experiences they're not old enough to use."

Meta's answer has been to infer age instead of trusting the date entered at sign-up. "We use AI technology to find people we suspect to be teens, even if they use an adult birthday, and place them in Teen Account protections," the post says. The company says that since first introducing the technology it has been "placing more teens in age-appropriate protections" and is "encouraged by the results." Based on that, "today we're expanding this technology globally."

The scope is narrower than the word globally suggests. Today's step covers Instagram. On Facebook, Meta says it is "continuing to roll this technology out," without naming the countries where it is or is not live. For Threads, the company says that "later this year, we'll begin launching similar technology globally," which, given the post's date of November 9, 2026, leaves about seven weeks. Messenger, which also carries Teen Accounts, is not mentioned in connection with the age technology.

From four countries to all of Instagram

The expansion completes a schedule Meta set out six months ago. On May 5, 2026, the company announced AI visual analysis that looks for cues such as height and bone structure in photos and videos to find children under 13, together with a wider rollout of the age prediction system aimed at 13- to 17-year-olds who list adult birthdays. That prediction system had first run in the United States, Australia, Canada and the United Kingdom. In May, Meta extended it to Instagram in all 27 European Union member states and Brazil, and to Facebook in the United States, with Facebook in the UK and EU due in June 2026 and global Instagram coverage planned for the rest of the year. Today's post is that last step.

Teen Accounts themselves are a little over two years old. Meta introduced them on Instagram in September 2024, making accounts for users under 16 private by default, limiting messages to followers and existing connections, and requiring parental permission before under-16s could loosen the settings. Meta said then that it would test age detection in the US in early 2025, and today's post links back to an April 2025 announcement as the starting point of the technology. That same month, Meta brought Teen Accounts to Facebook and Messenger and said at least 54 million Teen Accounts were active worldwide. In June 2026, it extended its 13+ content settings for teens globally across all three apps.

What Meta has not published

The post contains no count of accounts moved by the age technology so far, no estimate of how many will move now and no error rate. "Millions more" is the only quantity offered. PPC Land noted in May that the earlier announcement also gave no figure for accounts moved under the new measures.

That gap matters because the system runs on inference, and inference fails in two directions. An adult misread as a teen loses features. A teen misread as an adult keeps none of the protections. Meta does not say which signals its model uses, how someone flagged by mistake can contest the result, or how long a correction takes. The in-feed video prepared for US parents sums up the process in one line: "We estimate age, apply protections, and verify age when needed." When verification is needed, the post does not say.

The images published with the post show what a reclassified user sees. A banner at the top of the Instagram feed reads: "We've updated your safety settings. We occasionally update settings to improve your experience. We've added built-in protections to limit unwanted interactions and more. You can change these settings anytime." A "Review updated settings" button leads to three categories: "Who can interact with you," "What you see" and "Time management."

Nothing in that mock-up tells the user that the change follows an estimate of their age. Its line that the settings can be changed "anytime" also sits uneasily beside the rule, in place since the 2024 launch, that users under 16 need a parent's permission to make Teen Account settings less strict. Whether the text shown to a 14-year-old differs from the one in the image, the post does not explain.

The attorneys general agreement behind the rollout

Meta frames the global expansion as part of its settlement with US states. That settlement carries specific obligations on age, and they go further than the post describes.

According to PPC Land's reporting on the consent judgment, Chief US District Judge Yvonne Gonzalez Rogers signed it on August 26, 2026, and it took effect on August 27. It requires Meta to adopt an age assurance framework within one year, with ceilings on false positives in its under-18 age determinations: 10% for 16- and 17-year-olds and 3% for 13- to 15-year-olds when Meta uses commercial methods, and 14% and 7% when it uses its own proprietary methods, tightening to 10% and 5% within two years. Accounts that have not been assessed after 14 days are to be treated as teen accounts. For children under 13, Meta must maintain detection methods within six months, which falls in late February 2027, build a prototype under-13 model within one year and run it on all accounts within two years. The judgment also requires Meta to incorporate age signals from operating systems and app stores run by Apple and Google.

Today's post covers the same ground in broader terms. "As part of our agreement with 52 bipartisan attorneys general across US states, territories, and the District of Columbia, we're investing in even stronger technology to proactively catch accounts that may belong to children under 13," Meta writes. "We're also strengthening the technology we use to identify accounts that may belong to teens between the ages of 13 and 17 so we can ensure those accounts are placed in experiences designed for teens, even if they give us an adult birthday." The error-rate ceilings and deadlines go unmentioned.

The count of parties differs from the court record, too. PPC Land found that the settlement's definitions list 51 jurisdictions and omit Texas, which appears only in an exhibit as an eligible state. Meta has called the deal an agreement with 52 attorneys general since August and does so again today. The sums are large: Meta took a $9.89 billion charge in the third quarter of 2026 tied to the agreement, which its October 21 results release described as providing for payments of up to approximately $18.59 billion. PPC Land's own total from the court filing came to about $17.21 billion.

A video for US parents

Under the heading "Our Continued Partnership With Parents," Meta says that "this month, parents in the United States will begin to see an in-feed video on Instagram explaining our approach to ensuring their teens are in age-appropriate protections." The video links to an article by Larry Magid, described in the post as an "internet safety advocate and President and CEO of ConnectSafely," setting out what Meta calls its "multi-layered age assurance approach." According to the post, that approach "includes keeping children under 13 off our platforms, providing age appropriate experiences for teens, and informing parents about ways they can partner with us to ensure their teens provide the correct age online." The article sits on Meta's Family Center site and "is available for all parents worldwide to read."

In the image Meta published, the video runs in the feed under an "Instagram - Featured for you" label, with a caption that opens: "Instagram doesn't rely on one safeguard." This is the second US parent campaign in six months; the May 5 announcement also included a notification program for parents in the United States on Facebook and Instagram. How Instagram decides which US users are parents, the post does not say.

Apple's age-range signal in Australia

The third change is confined to one country. "In Australia, we've begun using Apple's age-range signal as another method to help understand age," Meta writes, before describing the process: "When a parent sets up a smartphone for their teen, they have the option for the app store to confirm the age of the person using the device and assign them an age range. Then, each time the teen downloads an app, Apple passes that age range to the developer - Meta, for example - confirming whether the teen meets the app's age requirement."

How Apple's signal works

Apple's mechanism is the Declared Age Range framework, introduced with iOS 26. It returns an age band, not a birthdate. The developer sets up to three age gates, which produce up to four ranges: gates at 13, 15 and 18, for example, return under 13, 13 to 14, 15 to 17, or 18 and over. The response states whether the range was declared by the user or by a guardian and, from iOS 26.2 in certain regions, whether it was checked by payment, by government ID or by another method. A separate field shows whether Screen Time or other parental controls are switched on. A changed band is not returned until the anniversary of the original declaration. Guardian declarations come from parents who manage a child's Apple account through Family Sharing.

Apple's age tools were already active in Australia. On February 24, 2026, Apple began blocking 18+ apps for adults who had not confirmed their age in Australia, Brazil and Singapore, according to PPC Land's explainer. The country's ban on social media accounts for under-16s took effect on December 10, 2025, with civil penalties of up to 150,000 penalty units, about A$49.5 million. In August, Meta said it had removed access from 756,000 accounts it assessed as belonging to under-16s between December 1, 2025 and June 30, 2026, 462,000 of them on Instagram and 294,000 on Facebook, and that it had raised its Apple App Store age rating in Australia to 16+. The same report carried Meta's argument for a single age signal delivered at the operating system or app store level, the position today's post repeats.

Voluntary in Australia, an obligation in the US

Meta stresses that it is acting by choice. "We're voluntarily using this signal, but as long as collecting it and sharing it remains optional for app stores, and using it remains optional for developers, teens will receive inconsistent protections across the many apps they use," the post says. The company adds that it has "long supported legislation requiring age verification and obtaining parental consent on the app store," and argues that "one verified signal from people's devices, shared consistently with all apps, is far more effective and privacy-preserving than asking every platform to verify age individually." Joel Kaplan, Meta's chief global affairs officer, made the same case in May.

The US position differs in two ways. First, according to PPC Land's reporting, the consent judgment already requires Meta to incorporate age signals from Apple's and Google's operating systems and app stores. Second, several states already require app stores to collect age and parental consent. Apple activated Texas SB 2420 rules for new Texas accounts in June 2026, requiring parental consent before minors download apps, after a court injunction had paused them in December 2025. PPC Land's explainer says Apple applies the same treatment in Utah from May 6, 2026 and in Louisiana from July 1, 2026. Today's post names only Australia and does not say whether Meta reads Apple's signal anywhere else.

One passage looks further ahead. "And the work developers like Meta do to estimate age can be used as a backstop to help correct mistakes and feed them back to the app store for correction across all the apps a teen uses," the post states. That describes age data flowing in the opposite direction: from Meta's own estimates, through the app store, to every other app on the device. The post does not say what data would be shared, whether the user would be told, or whether Apple has agreed to accept corrections from developers.

Laws that set ages but not methods

The post ends with a complaint about regulation. "Over the years, countries around the world have rapidly introduced new laws governing how to keep teens safe online," Meta writes. "While well-intentioned, many of these laws don't explain how technology companies should determine a user's age. Some require invasive age checks for everyone, while others offer no guidance at all - leaving companies to figure out how to understand age without unnecessarily collecting sensitive data, like government IDs."

No law is named. PPC Land's recent coverage shows how far the approaches diverge. Turkey barred social media accounts for under-15s from November 1, 2026 and routes verification through a token from its e-Government Portal that tells a platform only whether the user is over 15 or over 18; existing accounts must be checked by March 31, 2027. The UK government announced an under-16 ban on June 15, 2026 that relies on highly effective age assurance, with protections expected in spring 2027. The European Commission's EU KIDS Act proposal of September 17, 2026 would give providers six months to determine whether existing account holders are under 15 and would require them to disable accounts whose age cannot be established. In California, Meta sued on November 4, 2026 to block AB 1709, which bans personalized feeds and autoplay for under-16s from January 1, 2027.

Europe also supplies the clearest official account of what happens when a platform relies on the birthday users type in. On April 29, 2026, the European Commission issued preliminary findings under the Digital Services Act that Meta had failed to diligently identify, assess and mitigate risks to children under 13 on Instagram and Facebook. The Commission said children could enter a false birth date with no effective check, that reporting an under-13 account took up to seven clicks, and that roughly 10% to 12% of children under 13 use Instagram, Facebook or both. The findings are preliminary; a final decision could carry a fine of up to 6% of worldwide annual turnover.

"We've set a strong standard for age assurance, but unless other companies follow suit, teens remain inconsistently protected online, defeating the purpose of these laws," Meta writes. How strong? Without published error rates, the claim cannot be measured against anything in the post.

TikTok, YouTube and the name left out

The final sentence turns to competitors: "As Instagram is just one of the many apps teens use, we're also calling on our industry peers, TikTok and YouTube, to adopt these same practices to keep teens safe." The post links to an open letter Meta addressed to both companies in August.

The settlement names three companies, not two. According to PPC Land's analysis of its structure, Snap, TikTok and YouTube are the core industry members whose participation would trigger a second, stricter phase of time limits and release about $5.02 billion in contingent payments. Each would have to meet age assurance requirements no less restrictive than Meta's. Snap does not appear in today's post.

TikTok has already accepted similar terms in one state. In September, it agreed to a consent decree with Alabama that sets a two-hour default daily cap for teens and requires age estimation in 13-to-15 and 16-to-17 bands by September 25, 2027, with false-positive ceilings of 14% and 7% in the first year, tightening to 10% and 5%. Those are the same figures as Meta's targets for proprietary methods. Unless 40 attorneys general sign on, TikTok self-certifies its compliance based on internal testing. Google, which owns YouTube, began a limited rollout of machine learning age estimation for signed-in US users in July 2025. And on November 4, the same day as Meta, Google sued to block California's AB 1709 as applied to YouTube.

Why this matters for marketers and publishers

Meta's post does not mention advertising. Yet a system that reclassifies accounts by estimated age changes the make-up of the audience on one of the largest ad platforms in the world. Every account it moves into Teen Accounts was previously registered as an adult. The post does not say whether those accounts are also treated as under 18 for ad delivery, whether they drop out of adult audience definitions, or how the change would show up in reach estimates and delivery reports. Meta reported $63.06 billion in advertising revenue for the third quarter of 2026 and does not disclose how much comes from teen users.

Other platforms have tied age estimates directly to ads. Google's system, PPC Land reported, turns off ad personalization and blocks sensitive categories such as beauty and cosmetics or romantic content for users it judges likely to be minors. The EU's Digital Services Act bars profiling-based ads to users known to be minors, and Turkey's new rules bar the personal data of 15- to 17-year-olds from use in profiling or personalized commercial advertising. TikTok's Alabama decree bars data from users under 13 from being used for ad targeting and delivery. As Meta's estimates reach more countries, the number of Instagram accounts treated as teens, and any market-specific ad rules attached to that status, will rest on a model whose accuracy has not been published.

For app developers and publishers, the Australian move points to a different shift. Meta wants a single age signal from the device, shared with all apps, plus a channel for its own estimates to flow back into that signal. For now, Apple's framework is available to developers, but app store age collection is mandatory only where local law requires it, and each developer decides whether to read the signal. If Meta's argument prevails in more legislatures, the age band that tells Instagram to switch on teen protections would reach every ad-supported app on the same phone.

Timeline

Summary

Who: Meta, for Instagram, Facebook and Threads; teenagers aged 13 to 17 who registered with adult birthdays; parents in the United States, who will see an in-feed video linking to an article by Larry Magid of ConnectSafely; Apple, whose age-range signal Meta now uses in Australia; and TikTok and YouTube, which Meta calls on to follow.

What: Meta expanded the AI it uses to identify suspected teens to Instagram worldwide, moving those accounts into Teen Accounts regardless of the birthday entered. It said Facebook rollout continues and Threads will follow. It gave no count of affected accounts and no accuracy data.

When: Today, November 9, 2026. The US parent video starts this month, and the Threads rollout is due to begin later this year.

Where: Instagram globally, with the Apple signal used only in Australia and the parent video shown only in the United States.

Why: Meta says its teen protections depend on knowing users' ages and ties the expansion to its agreement with US attorneys general, which sets error-rate ceilings and deadlines for age assurance and under-13 detection. It also uses the announcement to argue that app stores should verify age for all apps and that TikTok and YouTube should adopt the same practices.