Consent and Digital Replicas: What Creators Should Know

Reading Time: 5 minutes

Published: January 9, 2026 | Last Updated: January 12, 2026

Digital replicas let you recreate someone’s face, voice, or performance using AI or visual effects. These tools open up creative options, but they also raise legal and ethical questions. You need permission that fits how the replica will be used.

This guide explains how to handle consent for digital replicas. It shows how permission works on real projects, what to document, and how to avoid common disputes. The article uses SAG-AFTRA terms as a practical reference, but the advice applies globally.

What Consent and Digital Replicas Mean

It’s easy to think that having footage means you also have permission to change or reuse it. But footage rights and replica rights are two different things. You need a clear definition before you move into post-production.

Consent for digital replicas means getting clear, written permission to use someone’s image, voice, or performance in a digital way. That can include altering a recorded performance, generating new lines, or creating versions for other languages, scenes, or platforms.

This applies across film, TV, commercials, online videos, and more. It includes replicas that are full-face, voice-only, or partial. It does not include fictional characters that don’t map to a real person.

Where Consent Questions Usually Appear

Consent issues often show up at certain moments in the production process. These points call for extra care and documentation.

• When you change what a person appears to say or do
• When you add performance that wasn’t recorded on set
• When you reuse a likeness in a new version, market, or campaign

What Counts as a Digital Replica in Practice

Digital replica sounds abstract until you look at how it’s used on set or in post. Once you name the type, you can ask for the right kind of permission.

Common Types You Will Encounter

Most replicas fall into a few common types. Each one affects how someone is seen or heard, so it’s important to be specific when asking for approval.

• Voice replica for ADR lines, language dubs, or technical fixes
• Face replacement for swaps, stunts, or background corrections
• De-aging or age change that shifts visual identity in a scene
• Performance synthesis that generates new actions or expressions from reference material
• Crowd replication that duplicates identifiable people at scale

Why Replicas Raise the Stakes

Substitutes have always been used in film. For example, a body double helps with stunts or sensitive scenes. But digital replicas keep the person’s identity front and center while the actions or words change. That makes consent easier to miss or misunderstand.

Why SAG-AFTRA Concepts Matter Outside the U.S.

Film projects often cross borders. Even if you’re not working under a U.S. union contract, SAG-AFTRA terms can help set clear expectations. They offer a detailed way to think about permission, reuse, and performer rights.

How SAG-AFTRA Defines a Digital Replica

SAG-AFTRA defines a digital replica as a computer-generated version of a performer. That version can be used to create new scenes, actions, or lines. If a tool can generate new performance, then the performer’s consent must cover that outcome directly.

What Informed and Specific Consent Looks Like

Vague contracts break when projects shift direction. Clear consent should answer key questions. What will be replicated? What will be created? Where will it appear? Who approves it?

Using These Standards in Non-Union Projects

You don’t have to be in a union to follow the same approach. Being clear about the scope, the use, and the approvals protects everyone involved. It also prevents misunderstandings if the project expands later.

Consent Scope: What to Lock Down Early

Many consent problems come from unclear limits. Writing down the scope from the start helps you avoid conflict if the project changes or grows.

Checklist of Scope Items

Here are the main items to define early. They help keep expectations aligned during production and post.

• Purpose: What the replica is for and what it can’t be used for
• What is replicated: Voice, face, body, performance, or a mix
• Context limits: Which scenes, lines, or versions it applies to
• Media, territory, and term: Where it runs, who sees it, and for how long
• Approvals: Who signs off on tests, versions, or changes
• Reuse rules: Whether it can be used in trailers, ads, social posts, or sequels
• Data handling: How it’s stored, who can access it, and when it’s deleted

How Standard Releases Compare

Regular release forms cover likeness and distribution, but they may not go far enough. You can start with a talent release or a model release, then add a separate section for digital replicas. Be specific about future reuse, approvals, and how data is stored or shared.

Workflow Example: AI Voice Replica for ADR

Voice replication comes up often in post-production. It can seem like a simple fix, but even one line can change meaning or performance credit. Here’s a sample process you can follow.

Step-by-Step Guide

This process fits into a regular post-production schedule. It keeps consent, quality control, and documentation in the same workflow.

1. Confirm the need: Write the exact lines you want to change and why (story, technical, client, or language version).
2. Define the scope: Is it just those lines? Can it be used in trailers or social edits?
3. Get permission in writing: Include details on voice replication, use, approvals, and term. Add an AI clause if needed.
4. Plan the source recording: Get clean voice samples. Log the date, setup, and location for reference.
5. Run a test and get approval: Share a short sample. If required, have the performer or rep sign off.
6. Only generate approved lines: Keep a record of tools, prompts, and who created the output.
7. Check meaning and fit: Make sure the new line matches character and scene tone.
8. Control access and archive: Store files securely. Set a deletion date if agreed in advance.

Managing AI Output Risks

AI-generated voices can fail in subtle ways. You may hear tone shifts, strange timing, or words that don’t quite fit the character. See FilmDaft’s guide to AI failure modes for more testing tips.

Common Mistakes and Misunderstandings

Many problems start with small assumptions. Calling them out makes them easier to avoid.

Footage Rights vs Replica Rights

Just because you can use recorded footage doesn’t mean you can generate new scenes or voice lines. Treat digital replication as a separate step that needs its own approval.

Tiny Changes Can Shift Meaning

Even one new line can change how a character feels, acts, or is judged. That matters to both the performer and the viewer. See our breakdown of suspension of disbelief to understand why small cues can break trust.

Realism Isn’t Always Safer

Hyper-realistic replicas can feel wrong even when they look polished. The uncanny valley effect can pull viewers out of the scene. The more realistic your replica, the more care you need to take with consent and context.

Context Varies by Country and Client

Consent rules depend on where you’re working. Some countries protect likeness or performance under personality or privacy rights. Distributors and clients may also add their own standards. Always check local law if the project involves known talent, sensitive themes, or wide exposure.

Summing Up

Digital replicas change how performance works. They also change what permission means. You need consent that is specific, written, and tied to use. SAG-AFTRA terms can help you think clearly, even if you’re not working in the U.S.

A clear workflow makes it easier to stay organized. Define the need, define the scope, document the approval, run a test, and store assets securely. These steps protect the performer, reduce production risk, and help the audience stay connected to the story.