Rightis

Integrate a service

Make your AI service check the rights before it uses anyone's face or voice.

With Rightis attached, your service asks whether the person is registered and what is permitted before generating, applies the contract's policy to every line, and records allowed and blocked generations alike. Connect over REST, OpenAPI or MCP, whichever suits you.

Why use Rightis

Rights check

Whether this person is registered and which rights may be used under which terms.

POST /api/v1/rights/check
Persona policy

Address style, relationship boundaries, knowledge cut-off, prohibited areas, safe lines. Versioned, with a webhook on change.

GET /api/v1/persona-licenses/{code}/policy
Approved assets

Contracted face and voice files via 10-minute signed URLs. Empty once the contract ends.

GET /api/v1/licenses/{code}/assets
Usage ledger

Send allowed and blocked generations alike. Replays are never stored twice.

POST /api/v1/usage
In-app registration

Your users register with Rightis without leaving your product. Rightis keeps the consent record.

POST /api/v1/rights-holders
Contract change events

License activation, expiry, revocation and settlement changes arrive as signed webhooks.

license.* settlement.* persona_policy.updated

Straight from AI agents

The same API as tools. Add the address under Connectors in Claude or ChatGPT, or as an app link in Gemini. The public lookup server (rightis.org/api/public-mcp/mcp) needs no auth; the partner server (rightis.org/api/mcp) takes your API key.

MCP server

Streamable HTTP. Pass the API key as a Bearer header and the key's scopes decide which tools open.

https://rightis.org/api/mcp
OpenAPI 3.1

Read as-is by custom GPT Actions, Gemini extensions and code generators.

https://rightis.org/api/v1/openapi.json
Webhooks

HMAC-SHA256 signature, per-event idempotency key, back-off retries.

X-BR-Signature: t=1757635200,v1=9f86d081884c7d65

What it looks like inside an AI tool

Before an AI tool makes someone's face or voice, it asks our public server first. Below is the real request and the real answer, exchanged just now.

Public MCP server/api/public-mcp/mcp

Can I train on this person's photos

tool call rights_lookup, rights_id BR-XV7E-8X7Z-ERB7
Checking

The line above is composed from that response. It is not a rehearsed sentence.

How to start

  1. 01
    Get a key

    Get a sandbox key right away in the developer console. A production key opens after review.

  2. 02
    Read policy and assets

    Fetch the policy and approved assets by license code and wire them into your generation pipeline.

  3. 03
    Send usage

    One event per generation. Settlement and audit follow from that.

Make your AI service check the rights before it uses anyone's face or voice. | Rightis