News
Meta Expands AI Age Estimation on Instagram Worldwide
Meta announced on November 9 that its AI systems for estimating users' ages are coming to Instagram worldwide. Accounts suspected of belonging to teenagers, including those created with an adult's date of birth, are to be moved into Teen Accounts.
Meta is expanding to Instagram worldwide the AI technology that assesses whether a user may be a teenager. If the system decides someone is 13 to 17 years old, it will move their account into Teen Accounts, even if an adult's date of birth was given at sign-up.
What Meta announced
According to Engadget, Meta is expanding its AI-based age-estimation systems so that more Instagram users end up in teen accounts. Until now the technology operated in the US, the European Union and Brazil. The Next Web reports that it was first deployed in the US, the UK, Canada and Australia, and in May it was extended to Brazil and the 27 EU member states.
Engadget says teen accounts block messages from strangers, make profiles and posts private by default, apply daily time limits in the app and include parental controls. Teen Accounts work across Instagram, Facebook and Messenger.
Parents and Apple's signal
In the United States, parents will start seeing a video in their Instagram feeds this month explaining how Meta places teenagers into age-appropriate settings and how parents can help confirm their child's real age.
In Australia, Meta has begun using Apple's age-range signal as an additional source of information. According to The Next Web, a parent can ask the App Store to verify their child's age, and Apple then shares an age range with apps such as Instagram when a teenager downloads them.
Meta has long supported legislation that would require app stores to verify users' ages and obtain parental consent. According to The Next Web, the company argues that a single verified age signal, shared by the device with all apps, would work better and protect privacy more effectively than separate checks on each platform.
The legal backdrop
Meta faces a growing number of lawsuits over harm to young users, The Next Web writes, citing Reuters. In August the company reached settlements with almost every US state and territory. In the agreement with 52 attorneys general, it committed to better identification of teenagers and children under 13.
Last month Meta also settled a case brought by a teenager in a Los Angeles trial over claims that Instagram and Facebook were designed to be addictive. Earlier this year, a jury in New Mexico found Meta liable for failing to protect children from predators and ordered it to pay $375 million in damages.
What remains unclear
Engadget corrected the original version of its story: according to Meta, analysis of bone structure in photos and mentions of things like school grades or birthdays are used to detect and remove users under 13, not to estimate teenagers' ages. As part of the August settlement with 52 attorneys general, Meta said, per Engadget, that it is "investing in even stronger technology to proactively catch accounts that may belong to children under 13". The company also wants to detect accounts belonging to 13- to 17-year-olds.
For users in Poland, the change mainly means that accounts in the EU have been covered by the system since May. What is new is the extension to the remaining countries and the move toward Threads. The sources do not describe exactly which signals the model uses or how long data from age estimation is stored.
