A couple problems with that are that people will disagree on what is inappropriate for children (e.g. drugs, nudity, sex, simulated violence, real violence, sim

A couple problems with that are that people will disagree on what is inappropriate for children (e.g. drugs, nudity, sex, simulated violence, real violence, simulated gore, real gore, foul language, educational material that's sexual in nature, hate speech, advertising), and parents may want unfiltered or differently filtered access for themselves while blocking it on a device/profile the child uses. I imagine a huge firewall list is probably also technically challenging for ISPs unless there's some convention for IPs (like even/odd LSB). But you can just have the site serve a content labels header. Specify what the content is, with a standard for topics which may be commonly considered inappropriate for children, and let parents configure controls at the OS level, which the browser should respect. This lets you even provide the info at the level of individual resources (so e.g. wikipedia could label images or pages with topics like nudity or violence, and parents could opt to allow/block topics). You could also have a coarse header indicating that the site does not want minors to access it, and that could be an option for the browser to block.