Meta plans to use AI to detect underage users through facial and body analysis

Meta will begin using artificial intelligence to analyse photos and videos for visual indicators that may suggest a user is under 13 and should not be on its platforms, the company announced on Tuesday. These indicators include characteristics such as a person’s height or bone structure.

“We want to be clear: this is not facial recognition,” Meta stated in a blog post. “Our AI looks at general themes and visual cues, for example, height or bone structure, to estimate someone’s general age; it does not identify the specific person in the image. By combining these visual insights with our analysis of text and interactions, we can significantly increase the number of underage accounts we identify and remove.”

The visual analysis system is currently active in select countries, with plans for a broader rollout underway.

Meta says the system is part of its wider efforts to prevent children under 13 from accessing its platforms, including Facebook and Instagram. These efforts also involve using AI to examine entire user profiles for contextual signals, such as references to birthday celebrations or mentions of school grades. The company evaluates these signals across multiple content types, including posts, comments, bios, captions, and more. Meta intends to extend this technology to additional features, including Instagram Live and Facebook Groups, in the future.

If Meta determines that an account is likely to belong to an underage user, it will deactivate the account. The user will then be required to verify their age via the platform’s age-verification process to avoid permanent deletion.

The announcement follows a recent legal ruling in New Mexico, where a jury ordered Meta to pay $375 million in civil penalties for allegedly misleading consumers about platform safety and exposing children to risks. The ruling also requires the company to implement significant changes to its platforms. In response, Meta has indicated it may shut down its services in the state.

This case is one of several legal challenges faced by Meta and other major technology companies regarding child safety concerns.

On the same day, Meta also announced an expansion of its “Teen Accounts” feature on Instagram to 27 countries across the European Union and Brazil. These accounts automatically apply stricter safety settings, such as limiting direct messages to approved contacts, filtering harmful comments, and setting profiles to private by default.

Additionally, Meta confirmed that it is introducing this technology to Facebook in the United States for the first time, with further rollouts planned in the United Kingdom and the European Union in June.