Meta expands Instagram’s AI age checks worldwide
Meta says its AI age-estimation system will place more Instagram users into Teen Account protections worldwide, but it has not published accuracy figures.
Meta is expanding its AI age-estimation technology worldwide to place more Instagram users into Teen Account protections, including people who entered an adult birthday. The company says the rollout will affect millions more teens and is designed to limit who can contact them, the content they see and the amount of time they spend on the app.1
What Meta is changing
Meta says its system looks for signals that an account may belong to a teenager, even when the user has supplied an adult date of birth. When it suspects that an account belongs to someone aged 13 to 17, the company says it can place the account into Instagram’s Teen Account experience. Those protections include restrictions on unwanted contact, age-appropriate content controls and time-management settings.1
The announcement is an expansion of a system Meta first introduced to identify teens who may have misrepresented their age. Meta says the technology is now being rolled out globally on Instagram. The company also says it is continuing a similar rollout on Facebook, while comparable technology is planned for Threads later in 2026.1
Meta has not published an accuracy rate, false-positive rate or false-negative rate for the age-estimation system in this announcement. It also has not provided a country-by-country schedule for the rollout. That means the announcement establishes Meta’s policy and deployment plan, but not how reliably the model identifies individual users in different regions or circumstances.
What parents and users will see
Meta says parents in the United States will begin receiving an Instagram notification this month explaining how the company places teens into age-appropriate protections and how parents can help correct an account’s age. The company is also publishing information for parents about its wider age-assurance approach.1
In Australia, Meta says it has started using Apple’s age-range signal as an additional way to understand whether someone falls within an app’s age requirements. Apple’s system can provide an age range when a parent sets up a smartphone for a teen, after which an app developer such as Meta can receive that signal when the teen downloads an app. Meta describes this as an additional method rather than a replacement for its own estimation technology.1
The distinction matters because Meta’s announcement does not say that every account will be checked in the same way. The company describes a combination of signals and says its own AI system is intended to catch accounts that may have been registered with an adult birthday. It does not disclose the model’s detailed inputs or explain how disputed classifications will be resolved in every market.
Why this is significant
Age assurance is becoming a central problem for social platforms. Services need to apply different protections to children and teenagers, but asking every user to submit an identity document would create its own privacy and security risks. Meta says it is trying to identify likely teen accounts without relying solely on government identification, while also calling for more consistent age signals across app stores and platforms.1
That remains Meta’s position, not an independent finding about the system’s effectiveness. The company says the technology has already placed more teens into age-appropriate protections and that the global expansion will improve safety. The available announcement does not include independent testing, an error breakdown or evidence comparing the system with other age-assurance methods.
Our opinion
The useful part of Meta’s announcement is that it acknowledges a practical problem: a birthday field is a weak safeguard when a user can simply enter a different year. Automatically applying stricter settings could close that gap for some accounts without forcing every family through an identity check. But the missing accuracy data is a serious limitation. A system that wrongly labels adults as teenagers can restrict legitimate users, while a system that misses teenagers can create false reassurance. Meta should publish clearer performance figures, explain how users can appeal an incorrect classification and state exactly where the rollout has reached. Until then, this is a substantial change in platform policy, but not proof that AI age estimation has solved age assurance.1