MENLO PARK, CA — Meta will strengthen privacy protections for children on Facebook and Instagram by first figuring out which users are children, a process expected to require the company to know significantly more about them.
Under an $18 billion settlement with U.S. states, Meta must improve age verification and safeguards for minors, including identifying users who may be lying about their age. Possible methods include analyzing friend networks, accepting reports of underage users, and temporarily using children’s data to train AI systems designed to detect children.
The breakthrough resolves a longstanding technology problem known as “you cannot protect information you are legally discouraged from collecting.”
“We take children’s privacy extremely seriously, which is why we need a more comprehensive understanding of their identities, relationships, behavior and approximate age,” a Meta spokesperson said while calmly adding several new columns to the database.
Federal privacy law restricts collection of certain personal information from children under 13, while increasingly aggressive age-verification requirements require platforms to determine whether the person whose information they shouldn’t collect is a child whose information they shouldn’t collect.
Meta said its AI systems will eventually become good enough at identifying children that users can once again enjoy the privacy of being extensively analyzed automatically.
Inspired by:
