U.S. Audiences Draw a Clear Line on AI in Film and TV: Tools Are Fine. Replacing Humans Is Not.

Luminate Survey Reveals Where American Viewers Accept Generative AI — and Where They Don't. The Findings Carry an Urgent Message for the K-Content Industry.

Generative AI can now write a screenplay, resurrect a deceased actor on screen, and redub a foreign-language series in flawless English. The technology is real, it is being deployed, and it is reshaping how film and television get made. But the people who actually watch that content have a more nuanced view than the industry's AI enthusiasm might suggest.

New survey data from market research firm Luminate — examining how U.S. movie and TV audiences feel about generative AI across specific production processes — delivers a verdict that cuts through the hype: Americans will accept AI as a backstage tool. They will not accept it as a replacement for human creators and performers.

For the Korean content industry, which has been among the fastest in the world to adopt AI production technologies, this data is both a warning and a roadmap.

The Headline Finding: Discomfort Outweighs Comfort, Across the Board

Luminate's survey asked U.S. film and TV viewers to rate their comfort level with generative AI being used across a range of specific production tasks — from sound effects and visual effects to scriptwriting, voice acting, and the creation of entirely synthetic performers.

The overall pattern was consistent: across most categories, the share of respondents saying they were "somewhat or very uncomfortable" exceeded those saying they were "somewhat or very comfortable." American audiences are not enthusiastically welcoming AI into their entertainment. They are watching it, cautiously, with their arms crossed.

But the data becomes more illuminating when you look at where the discomfort spikes — and where it doesn't.

Where AI Is Tolerated: The Technical Backstage

Viewers showed relatively higher acceptance — or at least an even split — for AI applications that enhance production without visibly displacing human talent.

Sound effects: AI-generated audio synchronized to on-screen action registered among the most accepted use cases. The logic is intuitive: audiences rarely think about how a punch sounds or how rain is recorded. When AI works invisibly and technically, resistance drops.

Age adjustment and high-quality VFX: Using AI to make an actor appear younger or older received comparatively warmer responses. Hollywood has been doing this for years; AI is simply making it more precise and affordable. Familiarity breeds tolerance.

AI dubbing: Perhaps the most strategically significant finding for the global content industry — AI-powered dubbing that re-voices foreign-language content in a way that matches the original actor's lip movements and vocal character was accepted at relatively higher rates. Viewers appear willing to accept AI when it preserves the actor's identity while removing the language barrier. This has direct implications for K-drama and K-variety distribution in non-Korean markets. The technology that makes a Korean actor "speak" English without losing their performance is, at least for now, within the acceptable zone.

Where AI Is Rejected: The Creative Core

The survey results shift sharply when AI moves from backstage technical work into what might be called the creative interior of a production.

Scriptwriting and scenario development: When AI writes the story, designs the dialogue, and structures the narrative arc, viewer discomfort rises significantly above comfort. The gap is not close. Americans appear to hold a firm belief that storytelling is a fundamentally human act, and they are not yet ready to surrender it to an algorithm — regardless of how sophisticated that algorithm becomes.

Narration and voiceover: AI providing the narrating voice for a documentary or the lead voice in an animated series also registered clear discomfort. The human voice carries emotion, imperfection, and identity in ways that audiences perceive as non-transferable. When AI takes the microphone in a creative capacity, rather than a technical one, viewers notice — and they object.

New character design and visual creative development: Even in animation and illustration, where AI is already widely used in the industry for concept sketching and ideation, viewer sentiment was divided and often leaned uncomfortable. Audiences expressed a vague but real anxiety: that characters and fictional worlds generated by AI carry a flattened, manufactured quality — what might be called an "AI stamp" — that dilutes the specificity of a creator's vision.

Luminate's analysis frames this pattern as a psychological "safety distance": the deeper AI penetrates into the creative decisions of a production, the less comfortable audiences become. The relationship is not linear — it is a threshold effect. Cross a certain line, and tolerance collapses.

The Hardest No: Digital Humans and Synthetic Performers

The three items that generated the highest discomfort ratings in the entire survey all involved replacing human performers in ways that are visible, embodied, and existential:

• Digitally recreating a deceased actor to perform in new content
• Creating a digital likeness of a living actor — their face, body, and voice — to perform roles they did not consent to
• Casting a fully synthetic AI performer — a "person" who never existed — as a principal character

Respondents described these scenarios in a phrase that Luminate flagged as telling: "technically impressive, but deeply uncomfortable." The two feelings coexist. Viewers can recognize the sophistication of the technology and still feel that something essential has been violated.

Luminate's report identifies this cluster as a source of significant latent risk for the industry. The legal and ethical exposure is real: right of publicity, post-mortem personality rights, consent frameworks for likeness usage — these are not abstract concerns. They are already the subject of active litigation and union negotiation in the United States.

The Screen Actors Guild (SAG-AFTRA) and the Writers Guild of America (WGA) have both moved aggressively in recent collective bargaining cycles to define the boundaries of permissible AI use and to establish compensation structures for AI-related work. The viewer data from Luminate suggests these unions are not operating in a vacuum — they are codifying, in contractual language, a sentiment that their members' audiences already hold.

Three Implications for the Korean Content Industry

The Korean entertainment industry has embraced AI faster and more broadly than most. AI influencers, virtual idols, AI-powered dubbing and narration, AI-assisted editing and synthesis — these are not speculative future technologies in South Korea. They are production tools in active use today. That speed creates competitive advantage. It also creates exposure.

Luminate's findings about U.S. audience attitudes carry direct relevance because the United States remains the world's largest and most influential content market, and because K-content's global reach now makes Korean production decisions subject to international audience scrutiny. Here is what the data demands:

First: Build contractual standards before they are forced on you. Korean production companies, agencies, entertainment labels, and broadcasters need to establish clear, standardized contract language governing AI use of actors', singers', voice actors', and creators' likenesses, voices, and performances. This means specifying scope, compensation rates, secondary use restrictions, and consent withdrawal mechanisms. The industry that writes these standards first will lead the next decade of global co-production negotiations. The industry that waits will have standards written for it — by foreign unions, foreign regulators, or foreign courts.

Second: Establish transparency frameworks for AI disclosure. Luminate's data strongly implies that audiences are not simply uncomfortable with AI — they are uncomfortable with undisclosed AI. Viewers want to know what they are watching. What proportion of a performance was generated or enhanced by AI? Which scenes contain synthetic elements? This information should be disclosed — in end credits, in production metadata, in promotional materials. Korean content that leads on voluntary disclosure will build a credibility premium. Korean content that conceals AI involvement risks the kind of sudden trust collapse that is very hard to recover from in the streaming era.

Third: Invest in domestic audience research. Luminate's findings reflect U.S. viewer psychology. Korean audience attitudes toward AI in entertainment may differ — in some areas more permissive, in others more protective. Without systematic, longitudinal survey data on Korean audience sentiment across specific AI use cases, production companies are making high-stakes creative and investment decisions without knowing how their own primary audience will respond. That research gap should be closed as quickly as possible.

The Strategic Opportunity: "Most Advanced Technology, Most Human Content"

There is a version of this moment that works strongly in Korea's favor.

K-content built its global brand on a paradox: production values that rivaled Hollywood budgets, combined with emotional authenticity and cultural specificity that Hollywood rarely achieved. The human element — the performances, the writer's voice, the director's singular vision — was not incidental to K-content's success. It was the product.

If Korean content companies can credibly position themselves as the industry that deploys AI most aggressively in the technical backstage — while protecting human creativity and performer rights most rigorously at the creative core — they can occupy a space no other major content-producing nation has claimed.

"Fastest AI adoption, strongest human protections" is not a contradiction. It is a brand proposition. And according to Luminate's data, it is exactly the proposition that audiences — in the United States and, almost certainly, globally — are looking for.

The question facing the Korean content industry right now is not "How much AI can we use?"

It is: "Where will we promise never to replace a human being — and mean it?"

That line needs to be drawn now, before the technology makes it feel too late to draw at all.

Source: Luminate, Generative AI in Film & TV Production — U.S. Audience Perception Survey (2026)Analysis: K-EnterTech Hub | k-entertech.com