Profile Bios Moderation

The moderation problem

A bio is public, persistent, and indexed. Unlike an ephemeral chat message, it sits on a profile indefinitely and shows up in search and previews. That raises the stakes: hidden slurs, evasive spellings, and self-promotion of hateful content all need catching before they're permanently visible.

Why a word list isn't enough

Word lists miss leetspeak and homoglyph evasion ('h4te', Cyrillic look-alikes) that users deliberately use in persistent fields, and they over-trigger on ordinary self-description that happens to contain a flagged token.

How The Profanity API handles it

Use context `bio` with `strict` mode so single matches flag and the blocklist layer's homoglyph normalization catches evasive spellings — appropriate for text that will be public and permanent rather than fleeting.

const res = await fetch("https://api.theprofanityapi.com/v1/check", {
  method: "POST",
  headers: {
    "Authorization": "Bearer " + process.env.PROFANITY_API_KEY,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ text, context: "bio", mode: "strict" }),
});

const { flagged, intent, score } = await res.json();