I agree that client-side filters would be a great solution, but I see two issues there:

1) Not everybody would know how to do it 2) This creates a weakest link problem, where in a class of say 15 kids, just having one with a non-blocked device would allow for all to see.

I don't know what would be good solution, maybe something intermediate... for example, filtering at the ISP level and making it mandatory for them to inform and request the settings for all their customers? Just a form, so they can block it. But then, maybe I want to block porn for my underage kids but not for me or my partner.

1) It's up to the parents to decide whether they want to put in the effort to look up how to use the parental control settings for their own child.

2) It's up to the child to decide who they want to associate with in school and in society, and up to the parents to advise their child in their decision-making.

Presumably, the parents are the ones buying the child's device, so this can be done at the OS level. The parent creates a user account for the child and a password-protected admin account for themselves on the new device, and only allow the firewall settings to be changed by the admin account. We can even implement offline on-device neural network-based detection and filtering, and you decide what to filter.

If the child is old enough to work and buy their own device, then it's debatable whether they should be moderated at all.

The problem with filtering at beyond the device level is widespread censorship, surveillance, and the erosion of the freedoms of the common man. The systems being built for supposedly the safety of the children are much too powerful that I can't help but question their true purpose.