Skip to content

Consent & compliance

The Studio is built so that every asset can be traced to a consenting, licensed likeness. The checks run automatically — you don’t have to police them, but it helps to know why something is sometimes blocked.

If you generate as yourself, you first enable usage on your model by confirming two consents:

  • Voice replication — permission to generate speech in your voice.
  • Synthetic media — permission to generate talking-head video of your likeness.

Building a 3D model additionally requires biometric consent to a body scan; the raw capture is deleted once reconstruction completes. When you upload reference material (rather than record it live) you also confirm a short rights attestation that you own or have the rights to use it. See Building your model.

Before any generation, the Studio confirms:

  • the project and licence allow this use,
  • the actor’s consent is active for it, and
  • you can cover the token cost.

If any check fails, generation stops with a clear reason — no partial or unlicensed asset is ever created.

A composition is only as compliant as the clips inside it. When you render or promote, the Studio re-checks every source asset the composition references — across all projects. If consent was withdrawn for any one of them, the render is blocked until you remove that clip. This is what makes cross-project and multi-character work safe: pulling a clip in never launders its consent.

If an actor withdraws consent, their affected assets are marked revoked. Revoked assets can’t be duplicated, rendered into new work, or promoted. Existing production assets are handled per your licence terms.

Development assets carry a visible draft watermark, and every asset is prepared to carry a content-provenance manifest so its origin can be verified downstream. Promotion is the point at which a take becomes a clean, production-ready asset.