Tilly Norwood: The AI Actress Dividing Hollywood

A Screenscope Guest Entry by Arin

We have entered the age of AI, where tasks are being automated, AI agents are ordering takeout for us, and AI actresses are starring in Hollywood films. Yes, you read that right.

Tilly Norwood is a fully AI-generated character—not a real person—created by the European digital studio Xicoia. Presented as a synthetic “actress” and “AI actor,” Norwood has generated significant controversy within Hollywood, with the actors’ union SAG-AFTRA condemning the concept and stating she is not an actor but a computer program trained on human actors’ work. Despite the backlash, this doesn’t seem to be a trend that will slow down anytime soon. Eline Van der Velden, Norwood’s creator, plans to develop at least 40 more digital stars.

Tilly has sparked one of 2025’s biggest film industry controversies. But is she actually a threat to actors, or just the latest in a long line of technological innovations that the industry eventually adapted to? This post explores what AI actors actually are, why they exist, and what they might mean for the future of film.

This Isn’t Entirely New Territory

The idea of digital performers isn’t completely unprecedented. In 2001, Final Fantasy: The Spirits Within introduced Aki Ross, a photorealistic CGI character designed to be cinema’s first fully digital actress. The filmmakers even gave her a backstory and modeled her to appear in future films. The movie flopped, and Aki Ross’s acting career ended before it began. But the ambition was there—Hollywood has been flirting with this idea for decades.

The difference now? The technology has advanced dramatically, and it’s becoming accessible enough that smaller studios like Xicoia can create their own digital performers. Whether the results are convincing is another matter entirely.

How Does This Actually Work?

Tilly Norwood was created using generative AI trained on millions of images and videos. The technology analyzes patterns in human faces, movements, and expressions to generate a synthetic character that can be directed and manipulated for different scenes and performances.

But here’s where it gets complicated: if you wanted to “sign” Tilly Norwood for a project, who exactly are you contracting with? Not Tilly—she doesn’t exist. You’re really signing with Van der Velden as the creator, or perhaps Xicoia as the studio. This raises fascinating questions about authorship and creative control. Is Van der Velden more like a director, an animator, a puppeteer, or something entirely new? The legal and contractual frameworks for this simply don’t exist yet, which is partly why major talent agencies like WME have announced they won’t represent AI talent. The traditional actor-agent relationship doesn’t translate.

The Business Case: Why Would Anyone Want This?

Let’s be honest about the economics. Hollywood productions are expensive, and actor salaries—particularly for A-listers—can consume a significant portion of a film’s budget. From a pure cost perspective, AI actors offer potential savings. You don’t need to coordinate shooting schedules around multiple stars’ availability. You don’t pay residuals. You don’t worry about contract renegotiations or actors aging out of roles.

For independent filmmakers working with tight budgets, the appeal is obvious. If you could tell your story with AI actors at a fraction of the cost, that might mean the difference between making your film or not making it at all. Van der Velden frames her work as creating a new art form—a way to enhance creativity and enable projects that might not otherwise be financially viable.

The comparison to self-checkout kiosks or automated ordering systems is inevitable. Those technologies were adopted because they reduced labor costs, even if customer experience sometimes suffered. But is acting the same as cashiering? The question is whether audiences will accept the trade-off, or whether there’s something fundamentally different about performance that resists automation.

Some argue this could open up new possibilities entirely. Perhaps AI actors could enable filmmakers to visualize stories set in dangerous locations or historical periods without logistical constraints. Maybe they could create characters that would be impossible with human actors. Van der Velden insists she’s not trying to replace actors but to create something new—a different category of performance altogether, more akin to advanced animation than traditional acting.

Is There Actually Demand for This?

This is the key question, and the answer is complicated. Van der Velden claims there’s interest from brands and production companies. The technology exists, the cost benefits are clear on paper, and there are certainly executives who see the financial appeal.

But demand from producers isn’t the same as demand from audiences. Early reactions to Tilly Norwood’s work have been largely negative. Critics pointed to exaggerated mouth movements and uncanny valley effects that made the performances unsettling rather than convincing. High-profile actors including Emily Blunt, Sophie Turner, and Toni Collette expressed disapproval. Even Ryan Reynolds satirized the concept in a Mint Mobile advertisement. The backlash was swift and unified, which suggests the market Van der Velden envisions may not materialize as she hopes—at least not yet. The technology might improve, but audience acceptance is a separate challenge entirely.

What About the Actors?

SAG-AFTRA’s condemnation was unequivocal: they argue the technology uses performers’ work without permission or compensation and directly threatens to put actors out of work. This is a legitimate concern, particularly for working actors who don’t have the star power to guarantee their irreplaceability.

But it’s worth noting that every major technological shift in filmmaking has initially frightened workers. When synchronized sound arrived in the late 1920s, silent film actors worried they’d become obsolete. Some did lose their careers, but many adapted, and the industry evolved. When CGI became mainstream, there were concerns about replacing stunt performers and practical effects artists. Motion capture technology raised similar questions about whether it was “real” acting.

In each case, the industry adapted. New roles emerged, and human creativity remained central even as the tools changed. Maybe AI actors will follow this pattern—not replacing humans entirely, but creating a new category that coexists with traditional performance.

Or maybe this time is different. The scale and speed of AI development is unprecedented, and the technology’s ability to replicate (or attempt to replicate) the core creative act of performance could represent a fundamental shift rather than just new tools. The question is whether acting can be reduced to patterns and data, or whether there’s something essential about human consciousness and interpretation that can’t be automated.

Star actors with established fan bases are probably safe—audiences come to see them specifically, and that’s not easily replaceable. But what about working actors, background performers, and voice talent? A two-tiered system where AI handles the bulk of roles while only a small elite remains employable might be more concerning than total replacement.

The Quality Question

Right now, Tilly Norwood’s performances fall into the uncanny valley—close enough to human to be recognizable, but just off enough to be disturbing. The mouth moves too much, the eyes don’t quite convince, and something indefinable feels wrong.

But is that a fundamental limitation of the technology, or just growing pains? AI image and video generation has improved exponentially in just the past few years. What seems unconvincing now might be seamless in five years. Then again, maybe there’s something about human performance—the tiny improvisations, the emotional authenticity, the way an actor brings theirown history to a role—that can never be fully replicated by an algorithm, no matter how sophisticated.

And even if the technology becomes visually perfect, will audiences accept it? We watch films partly for the craft and skill of performance, for the connection to another human being interpreting a character. Would watching an AI actor feel the same, or would it always carry a subtle sense of artificiality, like talking to a very convincing chatbot instead of a person?

A New Genre or a Threat to Tradition?

Van der Velden calls this a new art form. Is she right? Animation has existed as its own category for decades, with its own Oscars and its own audience expectations. Perhaps AI-generated performances represent something similar—not better or worse than traditional acting, just different.

If that’s the case, we might eventually see separate categories at awards shows. An Oscar for Best AI Performance wouldn’t necessarily diminish Best Actor any more than Best Animated Feature diminishes Best Picture. They’d simply recognize different types of artistic achievement.

Or maybe the distinction won’t hold. Maybe AI performances will get good enough that they compete directly with human actors for the same roles, the same awards, the same audience attention. In that scenario, the question becomes whether we value the art of performance itself, or whether we’re equally satisfied with a convincing simulation of performance.

The film industry has always been about illusion and artifice. Actors aren’t really the characters they play. Sets aren’t real locations. CGI creates impossible visuals. In that context, is an AI actor really so different? Or does it cross a line by automating the one element that has always been definitively human—the performance itself?

So What Now?

Is this the future of Hollywood, a passing fad, or something in between? Will we look back on Tilly Norwood as a curiosity from the early days of AI, or as the beginning of a fundamental transformation in how films are made?

The technology exists and will continue to improve. The economic incentives are clear. But technology and economics don’t tell the whole story. Audiences have to accept it. Actors have to decide whether to fight it or adapt to it. The legal system has to figure out how to regulate it. And we all have to decide what we actually value in performance and whether that’s something we’re willing to compromise on.

The industry is watching Tilly Norwood carefully—and what happens next might tell us more about what we value in performance, and in film itself, than we expect. Do we watch movies to see flawless, optimized content, or to connect with human artistry and emotion? The answer to that question will shape not just the future of acting, but the future of cinema.

What do you think? Is there room for AI actors in Hollywood, or is this a line we shouldn’t cross? The conversation is just beginning.