I think a reasonable way to divide up responsibilities would be identify child-locked devices, not people. It should be trivial for a website to identify a device with a child lock turned on. Then it's up to parents to make sure their kids get a child-locked device. Manufacturers can make it easy to turn on, and society can help by not selling devices to kids without a child lock turned on.

It won't work against a determined teen (too many unlocked devices out there), but it doesn't have to work perfectly to change the culture that most kids have to deal with.

This is part of the very sensible Digital Age Assurance Act that is currently the law in California. It prohibits facial recognition, but requires each device to have a child lock. FOSS can implement this no problem.

What disconnects this discussion from reality, and maybe causes questions like “why don’t they just do it this way”, is the assumption that this is really only or even mainly about age verification and “protecting the children”.

The reason it’s not done “this way” or “that way” even when those are objectively better ways to achieve the stated goal, but rather in an unexplainable way broader way is because the goal is broader that that and age verification is just the tip of the spear. The rest of it is laying the groundwork for a framework to control the freedom on the internet by linking identity to speech and action.

Look at what solution is implemented to decide what problem is it supposed to fix, otherwise you’re just looking at the smoke and mirrors.

Not that every state and country is on board with this, but it’s getting a lot harder to maintain the pressure to keep these initiatives down. Every time they get pushed one step forward it’s that much harder to regain that ground.