THE MISDIAGNOSIS
Age is a proxy, not a mechanism. The documented harms — sleep loss, comparison spirals, compulsive use — trace to the recommendation engine's optimization for engagement, not to a user's birth year. A 14-year-old and a 40-year-old fed the same algorithmic loop exhibit similar behavioral patterns.
WHY THE ID MANDATE IS THE POINT
To verify minors, platforms must verify everyone — adults cannot be presumed adult without proof. Every user uploads a passport, driver's license, or biometric scan. The age ban is, in implementation, a universal identity-verification regime sold under a child-safety frame.
THE COUNTRIES MOVING FIRST
Three jurisdictions have legislated under-16 bans with platform-side enforcement: Australia (passed late 2024, effective 2025), Norway (announced 2024), and Indonesia (drafted 2025). None has paired the ban with rules on how recommendation algorithms rank content for the users who remain.
THE EU PATH NOT TAKEN
The Digital Services Act, in force since 2024, requires very large platforms to assess systemic risks from their recommender systems and offer users a non-profiling feed option. It is the only major regulation that targets the algorithm itself rather than the audience — and the only one industry lobbied hardest against.
THE COPPA PRECEDENT
The US Children's Online Privacy Protection Act (1998) restricted data collection on under-13s. Platforms responded not by building safer products for children but by setting their minimum age at 13 and looking away. Age gates produce age lies; they do not produce safety.
THE LEVER NOBODY PULLS
Three algorithmic changes have shown measurable harm reduction in internal experiments: chronological feeds, friction on infinite scroll, and downranking of engagement-bait. None requires knowing a user's age. None has been mandated by any national law.