There is a global legislative consensus hardening in real time: children should not be on social media. Australia fired first in late 2025. The UK just followed, with Prime Minister Starmer announcing bans on TikTok, YouTube, Instagram, Snapchat, and X for anyone under 16. A TechCrunch tracker shows the list of countries moving in this direction is no longer a quirk. It is a movement.
Facial Recognition as Civic Infrastructure
The uncomfortable detail buried in the policy rollout is how age gets verified. Roblox's VP of safety, Eliza Jacobs, told NBC News that ticking a box is not enough anymore. The company is deploying facial age estimation. This is not a parental concern feature. It is biometric infrastructure being quietly normalized as child protection. Meanwhile, Big Tech's last-ditch lobbying push on AI regulation in Washington is packaging its own child safety rhetoric as a reason to soften broader oversight. The child becomes the rhetorical pivot point for every regulatory argument, on every side.
What the Muddy Children Puzzle Knows
A 2026 paper on arXiv by Hans van Ditmarsch titled History of the Muddy Children Puzzle traces a decades-old logic problem about shared knowledge and ignorance. In the puzzle, children cannot see their own muddy faces. They only know what others reveal. Age verification online operates on precisely this epistemology: platforms cannot see who is behind the screen without external signal. The solution, in both the puzzle and the policy, requires forcing revelation. The question is what else gets revealed in the process. Kyle Chayka's analysis of algorithmic homogenization applies here too: when the feed becomes the gatekeeper, every act of access is also an act of profiling. Age walls do not just keep children out. They teach platforms how to read everyone else more precisely.