The takeaway: Meta has announced that it has begun using AI to detect underage users on Facebook and Instagram and delete their accounts in accordance with company policy. According to Meta’s official terms of service, users must be at least 13 years old to create an account on either platform.

In a blog post, the Menlo Park-based tech giant said it is developing an advanced AI system to scan photos and videos on Facebook and Instagram, analyzing users’ bone structure, height, and other visual cues to estimate their age, while insisting that it is “not facial recognition.” The AI will also analyze user profiles for contextual clues – such as birthday celebrations or posts about school grades – to identify underage users.

If the AI algorithm identifies a user as a child under the age of 13, the profile will be deactivated, and the user will have to verify their age using an ID or Yoti’s facial age estimation tools within a specified time period to prevent the account from being deleted. The same mechanism will be triggered if someone attempts to circumvent the protections by changing their age from under 18 to over 18.

Meta noted that some of the new AI-based age detection algorithms have already been rolled out globally, but certain advanced features – such as visual analysis – are currently being used in only a handful of countries. The company said it plans to implement the new safeguards globally in the future.

Meta originally began using AI to identify teen users last year. The company claims the AI system has successfully enrolled “hundreds of millions” of teens into protective Teen Accounts, even when they attempted to circumvent safeguards by falsely identifying themselves as adults.

The AI age detection system is currently being used on Instagram in the United States, the United Kingdom, Canada, and Australia, and Meta is now expanding it to Brazil and 27 EU countries. The company is also expanding the protections to Facebook in the US for the first time this month, followed by the UK and the EU in June. Meta plans to implement the technology on Instagram globally by the end of the year.

The announcement comes just days after a New Mexico jury found Meta liable for jeopardizing the safety of children on its platforms by failing to protect them from child predators. The company was subsequently ordered to pay $375 million in damages for violating the state’s Unfair Practices Act, which prohibits unfair, deceptive, and misleading business practices.