ISPs come with adult content disabled by default and someone has to opt in to it. Every major OS (Windows, Mac, iOS, android) ships with device level parental controls. Games consoles enforce these based on birth date. ISPs here also provide free network level filtering on top of that. All of this only matters if the parents don’t bypass them when asked.
If a kid is determined enough to get past Apple family controls and the network level filtering on their home network, they’ll have a VPN from a dodgy source in 15 minutes. The solution is to use the tools that are there right now, or accept that age verification is coming for everything.
These are unfortunately rather half-baked and should be improved. Which is exactly what could be mandated instead of invading everyone's privacy.
You are right though - the fact that those controls exist and are in place and the UK government isn’t enforcing that Apple Microsoft and Google provide better tools (which would actually achieve the aim) tells you that what they actually want is what they’re asking for - a VPN backdoor.
what about whitelists? this never comes up anymore. I can load profiles from the 'child safety council' if that's what I want, and should expect to cover some of the overhead in evaluating all the submitted links. particularly in an educational setting, part of the problem is kids playing games and hanging out on social instead of working.
it seems a lot more tractable than trying to classify everything and get everyone to play along. let 1000 different filters bloom.
what's fundamentally wrong with that approach?
In practice I don't think it's an issue. What I'm arguing for is the infra to facilitate self categorization and (likely) also a legal requirement limited to only a few specific categories. For example the government might mandate that porn, social media, and user generated content all be accurately tagged and provide legal definitions.
Nothing about what I describe would preclude additional layers of categorization such as (but not limited to) whitelists. In fact it should improve such efforts by providing a standardized method they can use for arbitrarily fine grained categorization that will be compatible with other software out of the box.
Note that my tagging proposition could be applied per network request. So if the service operator wants to it should facilitate filtering out (for example) a comment section without blocking access to the rest of the site.
and instead of burdening the isp the publisher of mango sorbet recipes with ticking off all the right schema boxes, this can all be enforced at the consumer.
all the rest of these approaches kind of assume that there is 'reasonable' and 'unreasonable' content, and that we all mostly agree on the difference. which I think is fundamentally fallacious. do you really think we can agree, as a species, what PG-13 should mean for the entire internet?