Meta plans to use a new AI tool to prevent teens from lying about their age on Instagram

Meta has shared information on how it plans to use AI to catch teenagers who lie about their age on Instagram. Early next year, Meta is planning to use an "adult classifier" tool, that would reportedly be able to identify users who are younger than 18. The tool will then automatically apply Instagram's more restrictive privacy settings to them.

Meta's director of product management for youth and social impact, Allison Hartnett, said that the software would look for indicators like the accounts a user follows and the content they interact with regularly. Even if the profile says someone is over 18, the software would ignore that.

Back in September, Meta began rolling out teen accounts. For those accounts, the social media platform would apply its strict privacy settings to kids younger than 16. For example, such accounts are automatically set to private, and they can't message people they don't know.

Teens cannot change these settings without approval from a parent.

Of course, when it comes to software tools determining whether or not a person is underage, we're inevitably asking ourselves how accurate would that be. Meta has not disclosed yet how accurate this adult classifier tool would be.


On top of that, for teens who attempt to manually change the age on their account, there would be a requirement to prove their identity. This could be done with an ID document or a video selfie to Yoti.

Meta has previously partnered with Yoti for age verification for Facebook dating. An algorithm is used to estimate a person's age based on facial features - after Yoti shares the estimate with Meta, they both delete the video.

This comes amid criticism of social media platforms and their effects on teens from regulators and lawmakers. Meta plans to use this tool to ensure that teens are safe on its platforms. TikTok and other companies are also facing pressure for their effects on teens and younger users in general.

I'm a fan of enforcing strict policies to protect children on the internet. But instead of having a multitude of software tools that may or may not be accurate in determining someone's age, I'd rather make social media platforms paid so these companies would stop using addictive tactics to keep you on there.

If they're not addictive as much, it would be pretty simple to regulate. But most social media platforms earn from advertising, so they want to keep you online as much as possible. And younger users have a difficult time resisting this.