Dispatch 07 / Written from 2035
Scenario confidence: high · Evidence cutoff: 12 August 2026
Identity Rights Become Production Infrastructure
The synthetic performer of 2035 is not a file. It is a revocable relationship among a person, a role, a model, a territory, and a moment.
AI-generated conceptual scenario art, not documentary or forecast evidence.
The thesis
Digital replicas will force identity permissions out of contract records and into production systems. Responsible studios will treat voice, face, movement, style claims, and biographical identity as scoped, time-bound capabilities whose use can be inspected, compensated, renewed, challenged, and revoked.
Evidence / 2026
Signals visible from here.
The U.S. Copyright Office’s digital-replica report identifies gaps in the existing patchwork of protections and recommends a federal right for all individuals against knowing distribution of unauthorized replicas. The issue extends beyond famous performers.
The EU AI Act requires disclosure for defined deep-fake uses while providing context-sensitive treatment for evidently artistic, fictional, satirical, and analogous works. C2PA supplies technical building blocks for recording assertions and transformations, but it does not decide whether consent was valid or a use was fair.
Production contracts commonly describe permitted uses in prose that cannot travel with a model, derivative, or distribution request. This creates a widening gap between legal intent and executable media operations.
Inference / 2035
A future history.
A performer grants a role credential rather than surrendering an unrestricted likeness asset. It specifies character, emotional range, prohibited contexts, languages, territories, duration, review thresholds, compensation events, training boundaries, and posthumous policy.
When a director requests a synthetic pickup, the production system checks the credential, records the model and transformation, routes required review, attaches provenance, and calculates the contracted participation. A revoked or expired permission cannot silently continue through a cached model endpoint.
Identity stewardship becomes a creative service. Performers maintain authorized expressive models that preserve nuance across languages and access needs while retaining the right to refuse contexts that would damage personal or cultural integrity.
Interpretation
What changes for leaders and makers.
- Rights management must move closer to render time. A signed PDF in a document system cannot govern thousands of automated transformations without executable policy and human escalation.
- Casting expands rather than disappears when synthetic extension is treated as collaboration. The performer’s authored range, consent, and ongoing participation become part of the work’s value.
- Studios need an identity incident practice for unauthorized use, credential compromise, contested similarity, model leakage, and removal across derivative chains.
Recommendation / Now
Moves to make before the future hardens.
- 01Stop using perpetual, all-media synthetic-reuse language as a default. Separate capture, training, generation, transformation, distribution, and preservation permissions.
- 02Create a rights object that production tools can query while preserving the controlling human-readable agreement and a named escalation owner.
- 03Attach identity and transformation records to test assets, then verify that restrictions survive editing, transcoding, localization, and third-party delivery.
- 04Establish compensation and review rules before capturing material capable of generating a replica, including for employees and non-celebrity participants.
Open questions
What could change this conclusion.
Law will differ by territory, and technical policy will not resolve contested resemblance, parody, public interest, or collective cultural rights. Human adjudication remains essential.
Strict identity controls could privilege people with sophisticated representation while leaving others exposed. Industry infrastructure should support low-cost registration, challenge, and remedy.
Primary sources
The evidence beneath the scenario.
- Copyright and Artificial Intelligence, Part 1: Digital ReplicasUnited States Copyright Office · 2024-07
Finds gaps in protections against unauthorized digital replicas and recommends a federal right covering all individuals.
- Regulation (EU) 2024/1689: Artificial Intelligence ActEuropean Union · 2024-07
Establishes transparency duties for certain AI-generated or manipulated content, including deep fakes, with provisions for artistic works and editorial responsibility.
- C2PA Technical Specification 2.2Coalition for Content Provenance and Authenticity · Version 2.2
Defines signed, tamper-evident Content Credentials that bind provenance assertions to media while preserving the distinction between provenance and truth.
- Copyright and Artificial Intelligence, Part 2: CopyrightabilityUnited States Copyright Office · 2025-01
Concludes that generative-AI material can participate in copyrightable work when a human determines sufficient expressive elements, while prompts alone generally do not supply that authorship.