Instagram is testing new methods for users to verify their age, including an AI tool built by a third-party company, Yoti, that estimates your age simply by scanning your face.
Officially, you must be at least 13 years old to sign up for an Instagram account, but for years the company made little effort to enforce this rule. Until 2019, it didn’t even bother asking new users about their date of birth, let alone verify this information. However, after being talked about by privacy and child safety experts, Instagram has been introducing more and more age-verification features as well as methods of separating younger users from adults.
Currently, Instagram only asks users to verify their age when teens try to edit their date of birth to show them as 18 or older. To verify their age, users can enter pictures of different ID cardsand as of today, US users have two additional options: social vouching and AI estimation.
For the first method, social vouching, Instagram will ask three mutual followers of the user to confirm their age. The mutual followers must be over 18 years old and have three days to respond to Instagram’s request. The second method, AI estimation, involves sending a video selfie to a third-party company, Yoti, which uses machine learning to estimate a person’s age.
Yoti is a well-known player in the field of online age and identity verification, and its technology has been approved for use by the British government and German digital controllers† It uses various facial signals to estimate a target’s age. (Although the company itself says it doesn’t know what exactly these signals are† Wrinkles? A look of unspeakable sadness? It’s all in the mix.)
You can really try Yoti’s system on the internet here (the company says it doesn’t keep any data you share with it) and check out the accuracy percentages below. The numbers show, in years, the error rate of Yoti’s age estimates for different age groups, skin colors and genders.
The data shows that Yoti’s system is less accurate for female faces and dark-skinned people, and estimates for children under 24 to 2.5 years can vary. However, if the tool makes a rough estimate of a user’s age, its accuracy improves. A 2020 analysis of Yoti’s system by a third-party nonprofit found that it was 98.89 percent reliable in guessing whether people 18 or older were older or younger than 25. (That means for every 1,000 guesses the system made, there would be between 0 and 11 errors.)
It’s not clear how these numbers will translate for Instagram’s use case, or if the system is really secure enough for this purpose. For example, we were able to fool Yoti’s web demo by holding pictures in front of our webcam. And it’s not clear what could stop an underage user from doing the same, or enlisting the help of an older friend to pass the test.
This isn’t the only AI-powered tool Instagram uses to estimate users’ ages. Since last year, the company has been using automated systems to search for underage users. These tools scan information such as users’ birthday messages and the age of their friendship groups. For example, if a user claims to be 20, but posts about celebrating his 17th birthday, his account will be flagged and he may be required to verify his age. Instagram says it’s working on adding new data points to this system, and is even testing a language analysis tool that will determine whether a user is an adult or a teen based on how they write.