Meta is testing a new artificial‑intelligence system that scans visual cues—such as bone structure and height—to estimate whether a user is under 13 years old. The technology is designed to work alongside existing text‑analysis and behavior‑based tools that detect accounts violating the platform’s age‑restriction policies. When the AI flags a profile as potentially belonging to a minor, the account is temporarily disabled until the user can complete a formal age‑verification process.

The rollout is currently limited to a few pilot regions, with Meta planning a wider global deployment. The visual analysis will eventually be applied to a broader set of features, including Facebook Groups and Instagram Live, expanding the platform’s ability to protect younger users in real time.

The initiative comes amid growing regulatory scrutiny over child safety on social media. In parallel, Meta is extending its “Teen Accounts” framework, which enforces stricter privacy defaults, limits direct messaging, and makes accounts private by default. The framework is already active in 27 European Union member states and Brazil, with plans to bring comparable safeguards to the United States, the United Kingdom and additional EU markets later this year.