PrestoMint keeps the ledger that binds a verified person to the exact content they authorized — as cryptographic fingerprints, never the content itself. Platforms check it before they publish.
PrestoMint receives fingerprints and approval state — never images. It never becomes a content controller.
It accepts a verified-identity reference. ProntoID is the natural default; any trustworthy provider works.
Publishing runs a hash-membership check — a pure yes/no. No reading of contracts, no scope evaluation, no discretion.
A person verifies and approves a set of content a single time. That approval is minted into the ledger — and every future publish is checked against it automatically.
The reviewer proves they are a real, specific person through an identity provider’s own screens.
Identity layerThey view the actual content, streamed view-only and watermarked, straight from the platform.
PlatformThey sign off on an exact set — approving a subset is allowed. The signature blesses a manifest.
Release layerPrestoMint records the binding: this verified person, this fingerprint set, this version, this date.
PrestoMintAt publish, the platform asks one question — is every fingerprint approved? — and gets a yes or no.
PrestoMintImages only ever move between the reviewer and the platform. Everything PrestoMint and the identity provider handle is fingerprints, identity, and yes/no answers. That clean seam is the whole design.
PrestoMint maintains the authorization ledger and defines the interfaces around it. It deliberately owns neither of the implementations behind those interfaces — that restraint is what keeps it adoptable by any platform.
Three composable layers from Brooks & Keitt. They snap together for end-to-end content authorization — and each still stands on its own.
Verifies that a person is a real, specific human. General-purpose identity — the same check that powers KYC and age assurance.
PrestoMint + ProntoIDCaptures the signed release between the two parties and carries revocation. Consent, signed once, cleanly separated from identity.
Mints the approved fingerprint set against a verified identity, and answers the publish-time membership check. Content-blind, provider-agnostic.
Gate publication on a signed model release without ever reading the release or matching the person — a mechanical membership check that survives content being added, removed, or reshuffled across listings.
The model-release problem, made cryptographic. One review covers a whole shoot distributed across many products, with granular per-asset approval and revocation that propagates by membership.
Provable, revocable proof that an identifiable person consented to a specific use of their likeness — a purpose-built answer to a problem regulators and platforms are actively working through.
Send the fingerprints on a piece of content. Get back whether they sit inside an approved manifest for the depicted person. That’s the hot path.
# is this box cleared to publish? POST /v1/authorizations/check { "subject_ref": "pid_2f19c…", # verified identity "fingerprints": [ "sha256:9f4a…b1c7", "sha256:2d80…4e9a" ] } ← response { "authorized": true, "manifest_root": "a77f…03e1", "version": 3, "unmatched": [] }
We’re onboarding a first set of platforms alongside the reference implementation. If content authorization is on your roadmap, let’s talk.