Naven
Glossary
Content moderation
Deciding what a platform will and will not carry, and enforcing it.
Every platform moderates, including the ones that say they do not — refusing to remove something is itself a decision. What differs is who decides, against what written rule, how quickly, and whether the person affected is told why.
The usual shapes are reactive reporting, automated classification, human review, and some combination. Scale forces automation: a service with millions of messages a day cannot have a person read them, so it has software read them and people review the appeals.
Where Naven stands
Naven publishes its categories rather than describing them vaguely, screens before delivery rather than after report, and reviews its own refusals automatically — a second model checks every block and withdraws the ones that were wrong, because a warning that should not exist can suspend an account.