Too young to use Instagram? Meta’s AI Classifier could help catch teens lying about their age

Meta revealed that it is working on a new AI powered tool that will help determine if a person signing up for Instagram is a teen or an adult.

Several government agencies around the world are scrutinizing Meta, the parent company of Instagram, accusing the platform of posing risks to children’s physical and mental health from excessive use. While some countries already have some measures in place that prevent underage users from creating an account on social media platforms, the Australian government recently went as far as proposing a ban on social media for children under the age of 16.

In a blog post, Meta announced that it is working on a new AI-powered tool that could help catch teens who are lying about their age. While Meta-owned platforms like Instagram and Facebook require users to enter their age when signing up, many users lie about their age.

Meta says the AI model, dubbed Adult Classifier can “help determine whether someone is an adult (18 or over) or a teen (13-17)” and will automatically apply the appropriate privacy settings.

“Meta Introduces AI Tool to Detect Teen Users, Alters Privacy Settings and Messaging Restrictions”

The AI model detects whether a person is a teen or an adult by analyzing signals like profile information, account creation date, and their content and interactions with others. If the AI suspects that the user is under 18, it will mark their account and make it a teen account irrespective of their claimed age.

If the ‘adult classifier’ detects a teen using an account, Instagram will set the account to private and prevent the teen from messaging strangers. While Meta already had some measures in place that applied certain restrictions to teen accounts, with the new feature, the company altered its policy so that teens cannot change these settings without approval from their parents.

However, the new AI-powered ‘adult classifier’ tool’s accuracy remains unknown. In a statement to Bloomberg, Meta said that users falsely identified by the software can appeal by sharing government IDs or uploading a selfie.

Technology News Further Reading

More News