Meta Launches AI Tool to Strip Under‑13 Accounts Without Facial Scans

Context Meta, the parent company of Facebook and Instagram, announced a new age‑verification tool aimed at complying with U.S. child‑protection laws. The system will run across the company’s social platforms, scanning uploaded media and user‑generated content for age indicators.

Key Facts - The AI examines photos and videos for physical markers such as height and bone structure, then cross‑references textual clues like birthday posts or grade mentions. - Meta’s head of North American Safety explained that the tool can “look beyond someone admitting they are 12 years old” by aggregating signals from posts, comments and interactions to build an age profile. - The technology deliberately avoids facial‑recognition algorithms, a move that addresses privacy concerns while still providing a statistical estimate of a user’s age. - Accounts flagged as belonging to children under 13 are automatically deactivated, reducing the platform’s exposure to legal risk and protecting minors from unwanted data collection.

What It Means The rollout marks a shift from manual reporting to automated enforcement, potentially accelerating the removal of underage users. By relying on visual and textual cues rather than facial scans, Meta sidesteps a contentious area of biometric surveillance while still meeting regulatory expectations. Critics may question the accuracy of height‑based estimates, especially for users who post edited images or inaccurate personal details. However, the company argues that combining multiple data points improves reliability compared to single‑source verification.

Industry observers will watch how quickly the AI can identify and purge non‑compliant accounts and whether similar systems spread to other platforms. The next step will be monitoring false‑positive rates and any legal challenges that arise from automated deactivations.

What to watch next: Meta’s reporting on the tool’s effectiveness, user‑appeal processes for mistakenly removed accounts, and potential regulatory feedback on AI‑driven age verification.