I'm stuck on one ("Age assurance can be done"). My oldest is turning 14 in a couple of months. It's a weird age: I think several of them could fool an AI that they're over 18 (they'd fail instantly as soon as they spoke to an actual person), while half of them look the same as they did five years ago. I have little faith in AI as being a reliable way to tell age.

Personally, I have always looked a lot younger than my age. It sucked when I was a teenager/early 20's, but it has been awesome for the last couple of decades. I don't know that I'd have passed an AI check at 20. I certainly failed human checks: I was routinely ID'd everywhere I went until I was about 30.

Then point four goes and says:

We found a plethora of approaches that fit different use cases in different ways, but we did not find a single ubiquitous solution that would suit all use cases, nor did we find solutions that were guaranteed to be effective in all deployments.

Translation: this really is not simple at all and we shouldn't have opened the report saying it was doable straight-up.

It sounds like they're expecting the emergence of an age verification industry here. They list a pile of companies already who appear prepared to step into that role. You somehow identify to companyX that you are over 18, they provide you with some form of token and you can then use that token to be allowed to make accounts on sites. It's not clear who funds companyX's operations, however.

I'm not reading all 160-odd pages of this tonight, it's advisory at best. I'm still totally stuck on how you'd stop a 14 year old from installing an off-the-shelf Lemmy container into his/her homelab and started using it. And that assumes the report is right and do have a perfect way to verify everyone.