Meta is expanding its age-estimation systems with AI that analyzes photos and videos for visual cues that may indicate whether a user is under 13 or a teen using an inaccurate birthdate. The company says the technology looks at broad visual characteristics such as height or bone structure to estimate an age range, and does not identify the person in the image as facial recognition would.
The visual analysis will be combined with existing contextual signals from profiles, posts, comments, biographies and captions, including references to birthdays or school levels. If Meta identifies an account as potentially belonging to someone under 13, the account can be disabled and the user must complete an age verification process to avoid deletion. The company is also using AI models to help process user reports of underage accounts more consistently and quickly.
Meta is extending automated Teen Account protections as part of the same age-assurance push. Its technology for identifying likely teen users and placing them into stricter Instagram settings is expanding to 27 EU countries and Brazil, while Facebook users in the United States will also be moved into age-appropriate settings when the system determines they may be teens. Meta plans to bring the Facebook version to the United Kingdom and European Union in June.
Online platforms are facing new legal and commercial requirements to verify or estimate user ages without creating excessive privacy risks. U.S. state lawmakers have been advancing app-store age verification proposals, while age assurance vendors have been moving toward reusable and privacy-focused credentials through initiatives such as OpenAge.
Meta also repeated its preference for app stores and operating systems to verify age and parental consent centrally, rather than leaving each app to run its own process. The company says some of the new tools are rolling out globally, while more advanced visual analysis features are being introduced gradually in selected countries.
The rollout also separates age estimation from identity verification in Meta’s public framing. Users flagged by the system may still need to prove age through an ID document or third-party estimation tool, but the visual AI itself is described as a screening layer that assigns accounts to an age band rather than confirming a legal identity.
Sources: Meta
–
By the ID Tech Editorial Team
