Children Online: Pornography, Age Assurance, and Privacy

The strongest free-speech objection is that age assurance chills lawful reading, excludes adults without conventional documents, and supplies future governments with infrastructure that can be expanded beyond pornography. That risk is not speculative; identification at the point of reading changes behavior and creates valuable data. The answer is structural: narrow covered-service definitions, several proof methods, unlinkable threshold tokens, no stable identifier, deletion, private remedy, public testing, and expedited judicial review. A proposed system that cannot meet those conditions should fail even if its stated purpose is child protection.

The strongest child-safety objection is that privacy-preserving options and appeals make evasion easier, while a simple ID upload is available now. A leaky or easily bypassed system can impose the privacy cost without delivering the benefit. That strengthens the case for comparative testing, not maximum collection. The position holds: the lawful objective is reducing minor access, not building the largest possible identity file.

Adults bear friction, false rejection, and some loss of anonymity at the proof-provider layer. People without standard documents, disabled users, and people whose appearance falls near an age- estimation threshold face concentrated exclusion risk. Providers pay for verification, audits, security, deletion, and appeals. Smaller lawful publishers may face a greater relative burden. Parents still carry responsibility that no technical rule can automate. Children face residual exposure because every method can be evaded.

Multiple methods, accessibility, safe harbors, shared open standards, and a rapid appeal mitigate those burdens. The policy accepts some access friction as the price of enforcing a legitimate age line, while refusing identity retention and broad content scope that would make the privacy price larger than the child-safety gain.