Any AI system can train on your face right now. There’s no infrastructure to stop it.
Nicolas Cage is one of the most commonly deepfaked actors in the world — not because a studio authorized it, but because AI models can train on public footage with zero guardrails. ByteDance’s Seedance generated hyper-realistic Tom Cruise footage from public data alone. The studios you work with aren’t the problem — your contracts with them are solid. The threats are outside those relationships: AI companies training on public footage without authorization, synthetic replacements generated by entities that never hired anyone, and post-acquisition strangers who inherit performance data with no relationship to you. DIAP is the guardrail system the industry needs: an Identity Vault, module-level authorization controls, a real-time audit trail, and emergency revocation for unauthorized use. Free forever.
The unknown actor. No name recognition, no leverage, no recourse — until TrustMark.
30s
Voice sample needed to clone
0
Consent mechanisms available
Free
DIAP for talent — always
What DIAP Does For Talent & Creators
Identity Vault
Face, voice, and motion are registered as one-way embeddings in an Identity Vault. The raw biometric data is never stored. Agents and talent control who can access each module independently.
Emergency Revocation — Shared Safety Mechanism
Emergency revocation blocks unauthorized access — it never breaks valid licenses. If you licensed your identity to a studio for two years, that license is honored. It targets everything outside valid agreements: the unregistered app that never had authorization, the entity that generated content without a license, the organization that inherited your data with no relationship to you. Anyone in the authorized chain can trigger it — talent, agents, studios, or guild representatives — whoever detects unauthorized use first.
Real-Time Audit Trail
Every access request, token issuance, render receipt, and revocation is logged in the vault. Agents and talent see who requested what, when, and whether it was authorized — in real time.
ScriptModule — For Writer-Creators
For writer-creators, scripts deserve the same protection as face and voice. ScriptModule registers written work with a cryptographic fingerprint. Any AI system that wants to access, analyse, or train on registered work must request explicit authorization first. TRAINING_USE is a separate right — and it defaults to no. Every access is logged permanently via AuthorshipReceipts.
Sound IDs & Ultrasonic Watermarks
Every authorized audio output — dubbing, voice synthesis, any AI-generated audio using a registered voice — gets a unique Sound ID and an ultrasonic watermark embedded in an inaudible frequency range. Detectable by DIAP’s verification system. Audio verification works without the original file — meaning a voice can be tracked even if the original recording is never shared.
TRAINING_USE — Always Separate
Authorizing a studio to use your voice for dubbing does not authorize them to train a model on it. TRAINING_USE is a separate, explicit right that must be granted independently. It defaults to no. Always.
Agent & Manager Delegation
Agents, managers, and attorneys manage DIAP consent day-to-day — approving token requests, setting visibility preferences, reviewing audit logs, and triggering emergency revocation when they detect unauthorized use. This mirrors how talent representation already works. Agents can build or use branded management tools on top of DIAP’s API to manage their entire roster from one dashboard.
The Open Ecosystem — Third-Party Apps
Your talent agency, your union, or independent developers can build apps on top of DIAP’s infrastructure — a branded portal, a mobile app, a roster management tool. Your data stays in your DIAP vault. The consent logic follows DIAP’s rules. You can revoke any third-party app’s access instantly.
Free For Talent — Always
DIAP is free for individual talent. No subscription, no premium tier, no hidden costs. Identity protection is a right, not a product. Studios and platforms pay for API access. You never do.
How It Works For Talent & Creators
Register your identity modules
Self-service registration via phone camera or studio-quality capture. Face, voice, and motion are converted to one-way embeddings. The raw biometric data is discarded after extraction.
Set consent per module
Each identity module has independent authorization controls. Authorizing face rendering does not authorize voice cloning. Agents and talent set the rules for each module separately.
Monitor and intervene
Every request appears in the vault. Unauthorized attempts are blocked and logged. Agents or talent can revoke any authorization at any time with a single action.
Unknown App attempted FACE_3D access. Blocked.
Marcus Rivera registers face and voice modules in his Identity Vault
Studio A requests FACE_RENDER token — Marcus approves with 90-day terms
Unknown App attempts FACE_3D access — no token exists — request blocked automatically
Marcus sees the blocked request in his audit trail — one notification, full context
Marcus’s agent triggers emergency revocation on Unknown App — all future requests from that entity denied permanently
What Talent & Creators Get
- Identity Vault with independent module-level consent
- Emergency revocation — blocks unauthorized use without affecting valid licenses
- Real-time audit trail for every access request and render
- ScriptModule for writer-creators — cryptographic script protection
- Sound IDs & ultrasonic watermarks on all authorized audio outputs
- TRAINING_USE always separate, always explicit, defaults to no
- Self-service registration — no studio required
- Agent & manager delegation — your rep manages day-to-day and can trigger emergency revocation on threats
- Open ecosystem — third-party apps can build on DIAP, you control access
- Free for talent — always, no exceptions