The problem of security and privacy protection on social networks is a topic of great impact, especially when it comes to minors: this is why also Instagramin which numerous new features have already been introduced in recent months, is preparing to implement a system for verifying the age of users which will be based on the use of artificial intelligence.
The primary objective of Meta, therefore, is to prevent it from being possible for anyone to be able to circumvent the restrictions envisaged for the profiles of the youngest users of the app: with this in mind, the perfecting phase of a tool created ad hoc by the Menlo Park giant, named “Adult Classifier”which will be available from the beginning of 2025. As anticipated, a series of protection barriers are already present on Instagram: first of all, registration, as also happens with other social networks, should be prohibited for minors under 13 years of age and create already an important filter. For users aged between 13 and 17, however, the Meta developers have created strong limitations.
Currently, for example, accounts created by users under 16 are recognized as “Teen Accounts” and cataloged by default as “Private”, with all the restrictions applied to them: published posts can be viewed only and exclusively by followers who have received explicit approval from the profile owner, and it is impossible for adults to send messages to minors first . It could be an excellent starting point for protecting children and adolescents from sexual predators, if there wasn’t a small but big problem underlying the maintenance of these protections: everything collapses the moment the minor declares a false age when signing up on Instagram.
How can such a situation be avoided? Meta has decided to take action by creating a specific tool, on which it has been working for at least two years, which will have the main objective of finding out those who lie about their age: to create its “Adult Classifier”, Meta has exploited the potential of Amazon Bedrock AI models.
The tool will have a series of information available to understand whether the user is of age or not. Artificial intelligence, for example, will verify in real time the type of content you interact with, the accounts you follow or the content of some peculiar posts such as those relating to birthday greetings. Furthermore, it will be possible to understand, thanks to a specific database scanned by AI, whether the associated email or telephone number have already been used to create other profiles or whether the account is used on an electronic device previously connected with a “Teen Account”. When there is a suspicion that the person is a minor, and that a false age has therefore been declared, the Instagram system will apply the consequent restrictions: it will obviously be possible to forward appeal if you believe you have suffered them unfairly
In case of problems, the user will be asked to authenticate with a valid identity document or with a video selfie taken using Yoti,
age recognition platform already used by adult apps such as Facebook Dating: the file will be deleted after verification. The protection tool for minors will be available from the beginning of next year.