OPINION

Will AI Replace Actors and Singers? Mostly No

Why the performer stays human.

FAMILIAR LABS RESEARCHALL ARTICLES

ABSTRACT

Mostly no. AI will take over a large part of production: CGI, visual effects, likely the extras. The performers we pay attention to, the actor and the singer, will mostly stay human, and work with no human performer will stay a minority, even decades from now. We want human connection, and we feel less for a performance once we know no one gave it.

Performing is not production

PRODUCTION

  • CGI and visual effects, once the tools are good enough.
  • Likely the extras: the crowd, the passers-by.
  • Background music in an ad or a store.

PERFORMING

  • The actor whose face carries the scene.
  • The singer whose voice carries the song.
  • Whoever the audience came for.

AI will take on more and more of production, and that is fine. The performer is whoever holds the foreground: the actor, the singer, the guitarist whose riff carries the song. When the piano is central to a song, the pianist is one too; the line moves with our attention.

Why we want humans

The basis of humanity is loving and caring for one another, so we want to support humans, including the ones nobody knows yet. Hear a good song from someone new and you become a fan of the song and of the person who made it.

Younger audiences already love the smaller names because they are human.

NEW ARTISTSINDEPENDENTSCREATORSSTREAMERS

Empathy is not a passing bias

EMPATHY
WHAT PEOPLE WERE TOLD
The same replies, labelled human or AI (nine studies, over 6,000 people).
WHAT CHANGED
Rated less empathetic and less authentic once labelled AI.
COMPOSED
WHAT PEOPLE WERE TOLD
The same classical excerpts, credited to a human or an AI.
WHAT CHANGED
Liked less when credited to the AI.
PERFORMED
WHAT PEOPLE WERE TOLD
The same piano audio, credited to a pianist or an AI.
WHAT CHANGED
Rated lower in quality when credited to the AI.

In the piano study the audio was identical, and listeners still described differences in rhythm and dynamics (Shank and colleagues, 2022; Ansani and colleagues, 2025; Rubin and colleagues, Nature Human Behaviour, 2025). Some call that a bias that will fade. We think it is intrinsic to how we live, love and use language.

The label is coming either way: under the EU AI Act, realistic AI-made video and audio has to be disclosed from August 2, 2026.

Art made with AI, and our part

AI will make some truly great work, with high effort and high creativity: art made with AI, and welcome. Most of what people choose to support will still be performed by humans, and those performances deserve to reach everyone.

++

Understand every word. Keep the full familiar performance.

Dubbing is a craft done by skilled people, and it has always asked the audience for a trade: another voice, lips that no longer match. Familiar Labs carries the performance a person gave into another language and translates its meaning and its words. We never generate a person. We never replace an actor.

More: Song Dubbing, the opposite of AI music generation, Human to human, not AI to AI and our mission.

QUESTIONS

Will AI replace actors?

Mostly no. AI will take on much of production, CGI, visual effects and likely the extras, but the actors an audience comes to watch will mostly stay human.

Will AI replace singers?

Mostly no. Fully generated music has its place where nobody listens closely, an ad or a store. The singers people follow, new and independent ones included, will mostly stay human.

Does Familiar Labs generate actors or singers?

No. Familiar Labs never generates a person and never replaces an actor or a singer. It carries the performance a person gave into another language and translates its meaning and its words.

REFERENCES

  1. [1]Shank, Stefanik, Stuhlsatz, Kacirek, Belfi, AI composer bias: Listeners like music less when they think it was composed by an AI, Journal of Experimental Psychology: Applied, 2022
  2. [2]Ansani, Koehler, Giombini, Hämäläinen, Meng, Marini, Saarikallio, AI performer bias: Listeners like music less when they think it was performed by an AI, Empirical Studies of the Arts 43(2), 2025
  3. [3]Rubin, Perry, Goldenberg, Ong et al., Comparing the value of perceived human versus AI-generated empathy, Nature Human Behaviour, 2025 (Hebrew University summary)
  4. [4]Regulation (EU) 2024/1689, the AI Act, Article 50 (transparency), applies from August 2, 2026
  5. [5]Human to Human, Not AI to AI
  6. [6]The Opposite of AI Music Generation

Multimodal Translation

Familiar Labs builds generationally better dubbing with a Multimodal Localization Foundation Model (MLFM), the first foundation model for multimodal translation: one model decides what to localize across text, audio and visual, and shows you the choices. Your voice and lips, carried into new languages. They're a team of 7 with PhDs, backed by Samsung, RedBird, Gold House. Our mission

BACKED BY

  • Wheelhouse
  • Maximum Effort
  • Stanford
  • Yonsei
  • University of Waterloo