Meta's decision to use AI analysis of physical characteristics including height and bone structure to identify underage users is a significant technical and policy development that deserves careful scrutiny beyond the child safety framing Meta has used to announce it. The company has been explicit that the system is not facial recognition and does not identify specific individuals, which addresses the most serious privacy concern around biometric identification. What it does involve is automated analysis of physical developmental characteristics derived from photos and videos that users posted, an inference process that has its own privacy and accuracy implications that Meta has not fully addressed in its public announcement.

The accuracy question is particularly important given the consequences of a false positive determination. If Meta's system incorrectly identifies an adult user as potentially underage based on their physical appearance, that user faces account deactivation and an age verification burden to restore access. The populations most likely to be misclassified by a system trained to identify underage users based on height and bone structure may include petite adults, users with certain health conditions, or users whose physical presentation does not conform to the typical development curves the model was trained on. Meta has not disclosed its false positive rate or the demographic distribution of misclassifications, information that would be essential for evaluating whether the system achieves its child safety goals without creating disproportionate burdens for specific user groups. The broader context of the New Mexico judgment and the ongoing child safety lawsuit landscape makes clear why Meta is investing in visible enforcement mechanisms, but the governance around how AI physical characteristic analysis is applied to users who are adults deserves as much scrutiny as the child protection benefits the system is designed to deliver.