What Are AI Actors?

AI actors are computer-generated performers that fall into three distinct categories: digital replicas of real people, fully synthetic characters with no human origin, and AI-cloned voices used for dubbing or narration. Each works differently, carries different legal status, and displaces different parts of the industry. Understanding the distinctions matters if you're producing content in 2026.

Digital Replicas vs. Synthetic Performers

A digital replica (SAG-AFTRA's term) is an AI-generated likeness of a real performer, built from archival audio and visual material. The production workflow typically involves curating reference footage, using a live stand-in for motion capture, then overlaying the digital double's face and voice via AI compositing. Early landmark uses include the de-aged Mark Hamill in The Mandalorian and the Val Kilmer AI replica in As Deep as the Grave (2026), which was family-authorized given Kilmer's throat cancer.

A synthetic performer is different: a fully generated character with no real human as its base. The most prominent example is Tilly Norwood, unveiled by London-based Particle6 at the Zurich Film Festival in late 2025. Positioned as the first fully bookable AI lead actor, she sparked immediate backlash from talent agencies and unions. Her creator described her as "an artistic creation that took real humans time and care to develop" — a framing that did little to quiet the debate.

AI voice cloning is the third category: training models on a real actor's recorded voice to generate new dialogue. This practice has caused the most immediate harm to working performers, particularly in dubbing markets. In India, voice actor Ganessh Divekar — the Hindi voice of Pedro Pascal — discovered his voice had been used via AI without his consent. His response: "Earlier, I was a voice; now, I have to say I'm a human voice."

Where Union Protections Stand

SAG-AFTRA has negotiated AI provisions across several agreements since 2023.

The 2023 TV/Theatrical Strike Agreement established the baseline: informed consent and additional pay for digital replicas, a prohibition on using replicas to cross picket lines, and minimum payment rates for independently created digital doubles.

The 2025 Interactive Media Agreement added the ability for performers to suspend consent during a strike, covering video games and interactive platforms.

The 2026 TV/Theatrical Agreement, ratified in June 2026, went further. It requires companies to have an "articulable business reason" to scan a performer, commits producers to "a principle strongly favoring human performances," and limits synthetic performers to situations where they offer "significant additional value." SAG-AFTRA has also proposed a "Tilly Tax" — a fee charged every time a studio casts a synthetic performer instead of a human — designed to erase the cost advantage of AI talent.

On the voice side, a 2024 agreement with Replica Studios established an opt-in model giving SAG-AFTRA members control over licensing their voices. A separate pact with Narrativ covered AI digital voice replicas for digital advertising.

The Voice Acting Market in 2026

AI has already displaced human voice actors across a significant portion of standard narration work. The quality gap is real but closing: 2026 models add micro-pauses, breathing patterns, and context-aware intonation that earlier models lacked.

Audience data tells a more cautious story. According to the Audacy Innovation Tracker, human voice trust sits at 55% versus 23% for AI synthetic voices. Information recall shows a similar gap: 32.5% for human voices versus 14.3% for AI. Research consistently finds that trust ratings are highest when human voice is used.

The economics are paradoxical. Traditional voice-over rates are being undercut across the board. But actors willing to license their voices on opt-in terms are seeing licensing fees that can substantially exceed their traditional per-project rates. The market is bifurcating: commodity narration goes AI; premium character work and consent-based licensing goes to performers who negotiate well.

The dubbing crisis is geographically uneven. Mexico became the first country to ban AI dubbing following campaigns by voice actor collectives. Brazil, South Korea, and other countries have seen creative worker manifestos and new labor organizing. Fabio Azevedo, the Portuguese voice of Doctor Strange, has warned publicly that AI loses "Brazilian idiosyncrasies" — local emotional and cultural texture that trained human performers carry automatically.

What AI Actors Are Actually Replacing

The honest answer is: background roles first, lead performances last.

Background performers and extras are structurally at risk. So are stock-style commercial talent, brand spokespeople, and standard narration. These are high-volume, low-complexity performance categories where AI's consistency and cost advantages are clearest. Particle6 has claimed synthetic performers can cut talent-related costs significantly on a given project — factoring in fees, travel, insurance, and reshoots.

For independent creators, AI actors intersect directly with AI filmmaking workflows — particularly in micro-drama formats where casting real performers for every speaking role is cost-prohibitive. Understanding what an AI video generator actually does helps clarify how synthetic performers are composited into generated footage, which is a different process from replacing a human on a live set.

Lead dramatic performances are a different matter. Academic research published in Taylor & Francis journals in 2025 found that labeling content as featuring an AI actor reduces audience viewing intentions compared to identical content performed by humans. The 2026 Oscar season reinforced this: Sinners, Hamnet, and One Battle After Another (Best Picture winner) were all human-led storytelling films that dominated the conversation.

Actors bring improvisation, physical presence, and emotional nuance that current generative tools cannot replicate at lead-performance depth. Demi Moore, speaking at Cannes 2026, framed the situation directly: "AI is here...to fight it is to fight something that is a battle that we will lose." Her argument was for coexistence and negotiation, not replacement.

How Producers Are Responding

Some are moving fast. Disney signed a multi-year deal licensing characters to generative video tools. Doug Liman's Bitcoin thriller is being produced with Acme AI and FX. Kevin Hart produced a documentary that used a deepfake of Sam Altman as a central creative device.

Others are positioning AI as a support layer rather than a replacement. Critterz, which premiered at Cannes 2026, was described by AGC Studios as "human-led but AI-assisted."

Individual performers are also adapting. Matthew McConaughey filed trademark registrations to control his voice and likeness commercially. New roles are emerging on productions: AI supervisors, prompt directors, likeness rights managers, and synthetic performance coordinators.

Frequently Asked Questions

What exactly are AI actors?

AI actors are one of three things: digital replicas of real performers built from archival footage, fully synthetic characters generated without a human base, or AI-cloned voices used for dubbing and narration. Each category has different technical requirements, legal status, and industry impact.

Are AI actors cheaper than human actors?

For certain roles — background work, standard narration, commercial spokesperson spots — AI talent can reduce costs significantly. For lead dramatic roles, the quality difference and audience trust gap mean most productions still use human performers.

What protections do real actors have against AI actors?

SAG-AFTRA agreements now require informed consent and additional pay for digital replicas, restrict synthetic performer casting to cases with "significant additional value," and give performers the ability to suspend AI licensing during strikes. Mexico has banned AI dubbing outright. Performers can also register trademarks over their voice and likeness, as several high-profile actors have done.

Will AI actors replace human performers entirely?

Current evidence says no for lead performance work. Audience trust data and box office results both favor human-led content. The displacement is concentrated in commodity performance categories: extras, standard narration, dubbing in cost-sensitive markets. Emotionally complex, culturally specific, and long-form performance work remains human-dominated.

Keep Learning

For a broader look at how AI is reshaping production roles and workflows, see What Is AI Filmmaking? and How to Use AI in Filmmaking. If you want to understand how AI video models actually generate character motion from a single image, What Is an AI Video Generator? covers the mechanics in plain terms. The debate over AI voices vs. voice actors goes deeper on the dubbing and narration side of the conversation, and will AI replace actors? examines the long-term outlook for the performing profession.