She was built on work done without asking or paying the people who made it.
An AI character named Tilly Norwood sparked outrage from major actors like Sophie Turner and Cameron Cowperthwaite, who view it as disrespectful and a threat to human performers. The creator argues AI characters are creative works like animation, not replacements for actors, but industry professionals counter that AI was trained on their work without consent or compensation.
- Tilly Norwood, an AI-generated actress created by Particle6 founder Eline Van Der Velden, has been posting on Instagram since February 2025
- Major actors including Sophie Turner, Cameron Cowperthwaite, and Ralph Ineson publicly criticized the project
- Disney, Universal, and Warner Bros. have filed lawsuits against Midjourney for copyright infringement
- OpenAI's Sora now blocks videos generated in the style of living artists and offers public figures opt-out protections
Hollywood actors fiercely criticized an AI-generated actress named Tilly Norwood created by startup Particle6, citing concerns about unpaid use of their work to train AI models and potential job displacement.
In late September, a digital actress named Tilly Norwood began accumulating hundreds of angry comments on her Instagram account. She had been posting since February like any other aspiring Gen Z influencer—selfies, career updates, screen test announcements. But Tilly Norwood does not exist. She is a synthetic creation built by Eline Van Der Velden, founder of the AI startup Particle6, designed to demonstrate what artificial intelligence can do in film and television production.
The backlash erupted after the entertainment news outlet Deadline reported that talent agents were attempting to sign Tilly as an actress and that major studios had begun quietly adopting AI-generated content. In one recent post, the AI character boasted that in twenty seconds of work she had fought monsters, escaped explosions, sold a car, and nearly won an Oscar—all in a single day. The caption posed a challenge: "Find an actress who can do it all."
Actors responded with fury. Sophie Turner, known for Game of Thrones, left a terse comment: "No, thanks." Cameron Cowperthwaite, who appeared in Shameless and American Horror Story, called the project "incredibly thoughtless and frankly disturbing," expressing hope it would backfire entirely. Ralph Ineson, who starred in Nosferatu, offered a single-word dismissal. The anger was not abstract—it was personal, immediate, and rooted in a specific fear that had been building in Hollywood for years.
Van Der Velden responded by insisting that Tilly was not meant to replace human actors but rather to function as a creative work, a form of art comparable to animation, puppetry, or computer-generated imagery. She argued that AI characters should be evaluated on their own terms, as a distinct medium, rather than measured directly against human performers. The statement appeared on both her account and Tilly's Instagram page, an attempt to reframe the entire conversation.
But the defense rang hollow to many in the industry. Mara Wilson, who starred in Matilda and Mrs. Doubtfire, pointed out the core grievance in a comment: Van Der Velden did not create Tilly alone. Hundreds of real workers—photographers, cinematographers, and countless others—had their labor embedded in the training data that made the AI possible. Their work was used without consent and without compensation, then repurposed to generate a product that could replace them.
This dispute sits at the center of a larger reckoning in Hollywood. Writers and actors had already fought a major battle in 2023, striking for months over concerns about how studios could use artificial intelligence. The unions won protections limiting how major studios and streaming services could deploy AI technology. But those agreements cannot prevent independent creators or smaller companies from using AI tools trained on vast swaths of the internet to generate content that mimics human actors or recreates existing film scenes.
Major media companies have begun fighting back through the courts. Disney and Universal sued Midjourney in June, accusing the image and video generator of illegally training its AI on their materials and then producing unauthorized recreations of beloved characters like Bart Simpson and Wall-E. Warner Bros. filed a similar lawsuit against Midjourney weeks later. The legal strategy is clear: if AI companies want to use copyrighted material, they must ask permission and negotiate terms.
OpenAI, the company behind the widely used Sora video generator, announced new safeguards this week. The updated version of Sora, released alongside a standalone application, will now proactively block videos generated in the style of living artists and will offer public figures the option to prevent the technology from recreating their image. The company stated it is working with copyright holders to understand their preferences about how their content appears in the AI ecosystem. Yet the fundamental tension remains unresolved: the technology exists, it works, and the incentives to use it are enormous.
Citas Notables
She is not meant to replace a human actor, but rather functions as a creative work, comparable to animation or puppetry.— Eline Van Der Velden, Particle6 founder
You did not create this. Hundreds of real workers—photographers, cinematographers—did. You took their work and pretended it was yours.— Mara Wilson, actor (paraphrased from her Instagram comment)
La Conversación del Hearth Otra perspectiva de la historia
Why did this particular AI character spark such an immediate and visceral reaction? There are other AI projects out there.
Because Tilly looked real enough to be hired. She wasn't abstract or obviously synthetic. Agents were actually trying to sign her as an actress. That crossed a line from "interesting technology" to "this could take my job."
But the creator says Tilly is art, like animation. Why doesn't that argument work?
Because animation requires animators. Puppetry requires puppeteers. Those are human jobs. Tilly was built by training an AI on thousands of hours of film and performance—work done by actors, cinematographers, directors—without asking them or paying them. The creator didn't invent the technique; she borrowed everyone else's work.
So the real issue is consent and compensation, not the existence of AI itself?
Exactly. If Particle6 had licensed the training data, negotiated with actors' unions, and built Tilly transparently, the reaction might have been different. Instead, it felt like theft dressed up as innovation.
The studios are suing now. Will that actually stop this?
It might slow it down, but probably not stop it. The lawsuits are about copyright infringement—specific characters, specific scenes. But the underlying technology is already out there. What might actually matter is if the unions can negotiate stronger protections in future contracts, and if enough public pressure makes studios hesitant to use AI-generated talent.
What happens to actors in the meantime?
They wait and worry. The technology is improving fast. In five years, it might be indistinguishable from human performance. That's what keeps them awake at night.