IMO the issue is way deeper than the method of age verification because you cannot automate child safety (from what I’ve read Roblox and Youtube Kids are notable failures here), and tech platforms are going to pick between some half-assed attempt at such automation and entirely banning minors, at which point they find a way to bypass it or go somewhere else on the internet that is also unsafe for children. It’s all about deferring blame to someone else, not creating a society where people are making sure the kids are ok, and cannot possibly address the actual problem.

But if the idea is that people are too stupid to understand this and it’s necessary to come up with the least bad form of plausibly functional age verification to placate them, maybe the best bet would be something like those self-attesting OS verification laws but modified to be less shitty. Something like a more voluntary system where when you buy a device, there’s a prominent option to get a version of it that is for kids and has parental control software preinstalled as bloatware. This could (optionally!) broadcast a flag (only to preapproved apps and websites!) that lets the service block the user from seeing or interacting with certain stuff, and otherwise do parental control things. It could at least somewhat work since kids don’t have the ability to buy their own hardware. It could be open source and collectively developed and funded as a set of common standards that experts have thought about and agreed on.

It wouldn’t solve the biggest issues, but at least it might make it a little simpler for parents who are overwhelmed and need a simpler solution for handling the problem of the raw internet.