AI-created performer Tilly Norwood is set to lead “Misaligned,” a feature film being developed by the company that created her. The project will test more than the quality of synthetic video. It will test whether audiences will pay to watch a screen personality who has no life or career outside the software and people behind her.
XICOIA describes Norwood as its first digital star and says she can work across film, television, advertising, games and live virtual events. Its parent company, UK production studio Particle6, announced “Misaligned” in July.
The film is not yet a finished product or a scheduled theatrical release. Variety reported that it was in early development, with collaborators still being attached. Particle6 also said it would be a hybrid production involving writers, directors, editors and other film professionals working alongside AI specialists.
Norwood may appear alone on screen, but a synthetic performer is still the output of human decisions, computing systems and source material. The empty space where an actor would normally stand does not mean the production is workerless.
An AI performer is not a digital replica
Hollywood’s contracts distinguish between synthetic performers and digital replicas.
A digital replica recreates an identifiable real person’s face, voice or performance. A synthetic performer looks or sounds like a person but is not recognizable as one particular individual.
That difference affects consent and ownership, but it does not remove the employment issue. A synthetic character can still fill a role that might otherwise have gone to a human performer.
The latest SAG-AFTRA television and theatrical agreement requires producers covered by the contract to notify and bargain with the union before using synthetics. Producers also agreed to favor human performances and not use a synthetic in a human role unless it adds “significant additional value” compared with a human or a digital replica.
These provisions do not ban artificial characters. They make their use in union-covered productions a matter for negotiation rather than a simple technical choice.
The business case is ownership and scale
A successful artificial performer could be commercially attractive for reasons that have little to do with acting quality.
A company can define the character, control its public image and deploy it in several markets at once. XICOIA says its digital talent can appear in scripted productions, brand campaigns, games and virtual events, while speaking to different audiences without dropping out of character.
That makes a synthetic performer look less like a conventional employee and more like intellectual property that can front several products. The same character might star in a film, endorse a brand and interact with fans online under one coordinated ownership structure.
Particle6 also says its AI production methods can reduce costs by as much as 90% compared with traditional production. That figure is a company claim, not an independently established saving for a Hollywood feature film.
The comparison also depends on what is being made. An AI-generated short, advertising clip or digital background is not equivalent to a feature with a large cast, complex rights clearances and a theatrical distribution campaign. Particle6 itself says high-quality AI filmmaking still requires substantial human craft, judgment and time.
Cost savings may therefore arrive in particular tasks, such as creating some environments, localizing material or reducing the need for certain physical shoots. They should not be confused with eliminating most of a film’s budget.
Audiences are more accepting of AI behind the camera
For studios, the decisive constraint may be demand rather than production technology.
A YouGov survey of 5,382 British adults in October 2025 asked how an AI-generated actor would affect their willingness to watch a live-action film. Only 5% said they would be more likely to watch, while 60% said they would be less likely. Another 21% said it would make no difference.
Advertising produced a similarly difficult result. In July 2026, YouGov found that 76% of British adults who had seen AI-generated actors in advertisements disliked them. Sixty-eight percent said they were less convincing than human actors.
However, those findings do not mean viewers reject every use of AI in entertainment.
A peer-reviewed experiment involving 500 people, published in the Journal of Cultural Economics in 2025, found no discernible penalty when participants were told AI had helped create film synopses and make casting and director suggestions.
That study examined film development, not a synthetic actor delivering the performance. It supports a narrower conclusion: viewers may accept some AI involvement without accepting the replacement of the person on screen.
Deloitte’s 2026 Digital Media Trends research points to the same divide. Nearly 40% of entertainment fans said they would accept clearly labeled AI-created content across streaming, social media, music and games. About one-quarter were open to interacting with virtual AI personalities if they were interesting or informative.
Acceptance, in other words, appears to depend on the format, disclosure and the job the technology is doing.
Smaller roles may face pressure first
An artificial performer does not need to displace a major movie star to alter the labor market.
Producers may find it less risky to test synthetic characters in advertising, background work, short-form content or lower-budget productions before entrusting them with an expensive feature. Those uses could reduce some entry-level opportunities even if established stars remain in demand.
This remains a possible path, not a measured employment outcome. Union contracts, audience preferences and future laws will influence how far substitution goes.
There could also be an effect in the opposite direction. If synthetic faces become common in inexpensive content, well-known human performers may become more commercially distinctive. A star brings an established audience, publicity and a history of previous work. Those assets are difficult to manufacture alongside a digital face.
Who owns an artificial star?
XICOIA says its characters are original intellectual property and that it can develop commissioned characters under agreed ownership arrangements. The legal protection surrounding AI-generated material is less straightforward than that marketing language may suggest.
The US Copyright Office has concluded that generative AI output can be protected when it contains sufficient human authorship. Material determined by a machine is not made copyrightable simply because a person entered prompts.
A production may still contain copyrightable human contributions, including a script, editing decisions, music or the selection and arrangement of material. Names, logos and commercial identity may also involve trademark and contract rights. The exact protection will depend on what people created and how the finished work was assembled.
The Copyright Office has separately recommended federal legislation covering unauthorized digital replicas, finding gaps in the existing mixture of state and federal protections. That issue concerns copies of real people rather than an original synthetic character, but both questions sit inside the same commercial problem: a studio needs clear rights before it can reuse a screen identity across films, advertisements and markets.
“Misaligned” will not settle all of these questions. It may not even prove that audiences are ready to accept Norwood as an actor rather than a visual effect or animated character.
But it will test the part of the business case that cannot be solved by faster video models or cheaper computing. A production company can create a digital performer and own the character behind it. Turning that character into a star still requires people to care.
