Meta Uses AI to Detect Underage Users via Body Metrics

Meta is rolling out a new artificial intelligence system designed to identify and remove users under the age of 13 from Facebook and Instagram. The technology scans photos and videos for specific visual indicators, such as a person’s height or bone structure, to estimate their age.

Meta will use AI to analyze height and bone structure to identify if users are underage

The company announced the update on Tuesday, emphasizing that the tool is not a facial recognition system. According to Meta, the software focuses on general visual cues rather than identifying specific individuals, combining these physical insights with an analysis of user interactions and text patterns to flag prohibited accounts.

Expanding Detection Capabilities

Currently active in select countries, the system is part of a broader push to enforce age restrictions. Beyond physical analysis, Meta’s AI monitors profiles for contextual clues, such as references to school grades or posts about birthday celebrations. These signals are tracked across various app elements, including:

  • Bios and profile captions
  • Comments sections
  • Public and private posts

Meta intends to integrate this technology into additional features, including Facebook Groups and Instagram Live. If the system flags a profile as potentially underage, the account will be deactivated unless the user can complete an age verification process to prove they meet the minimum age requirements.

Legal Pressure and Platform Safeguards

The deployment of this technology follows significant legal challenges regarding child safety. A New Mexico jury recently ordered Meta to pay $375 million in civil penalties, citing that the company misled consumers about safety risks to minors. The ruling also mandated structural changes to the platforms, with Meta responding by threatening to withdraw its services from the state.

In parallel, Meta is expanding its “Teen Accounts” initiative, which enforces stricter privacy settings by default. This includes restricting direct messages to existing connections and filtering out harmful comments. This feature is currently expanding to 27 countries across the EU and Brazil, with plans to launch the technology on Facebook in the United States, followed by the U.K. and EU in June.

Deixe um comentário

O seu endereço de e-mail não será publicado. Campos obrigatórios são marcados com *