AI Everything Today

Showrunner wants to recreate lost footage from Orson Welles using a new generative AI

Showrunner, an AI startup, says it has built a generative model called FILM-1 and plans to use it to approximate lost scenes from Orson Welles’ 1942 film The Magnificent Ambersons. The company has invited AI VFX artist Tom Clive and filmmaker Brian Rose to help. Warner Bros. Discovery holds the rights to the film, and Showrunner has a prior history of releasing unauthorized AI-generated episodes based on existing intellectual property.

Showrunner describes the idea as a cultural restoration, and it says the project will not be monetized. The plan raises clear legal questions about copyright and rights holder consent, and ethical questions about recreating the images of deceased actors, authenticity, and how generative AI should be used on historical art.

What Showrunner says it will do and how FILM-1 works

The company proposes a hybrid workflow that mixes AI generation with real-world filming. FILM-1 is meant to produce keyframes, which are reference images that define a scene. Those AI-generated frames would guide live actors, and then face manipulation would map actors to look like the original performances. The model also uses archival set photos and other spatial context to approximate how lost scenes might have appeared.

Steps Showrunner plans to take

• Generate AI keyframes with FILM-1 to suggest composition, lighting, and camera framing.
• Use set photos and archival materials for spatial and period accuracy.
• Film actors performing the scenes with the AI keyframes as visual guides.
• Apply face-manipulation techniques to align the actors’ faces with historic likenesses.
• Combine AI-generated and live footage to create an approximation of the missing sequences.

Who is involved and what each brings

Showrunner is the startup driving the work. FILM-1 is its new generative model aimed at film-level outputs. Tom Clive, an AI VFX artist, will offer technical expertise in visual effects and face-manipulation workflows. Brian Rose, a filmmaker who previously attempted a restoration project for The Magnificent Ambersons, brings archival research and filmmaking experience. Warner Bros. Discovery owns the film rights, which is central to any legal path forward.

Why this matters to ordinary readers

This project touches on several questions many people will encounter as AI tools improve. These include who controls a cultural object, how we treat the images of people who are no longer alive, and what counts as authentic restoration versus a new creative work. The answers affect access to movies, how film history is taught, and what rules studios and platforms will set for AI re-creations.

Practical implications

• Access to cultural heritage: If allowed, AI-assisted restorations could make previously lost or damaged art more viewable.
• Copyright and control: Rights holders such as Warner Bros. Discovery decide what can be shown and distributed.
• Trust and authenticity: Viewers and historians will need clear labels and methods to know whether a scene is original, reconstructed, or partly synthetic.
• Personal likeness and consent: Families, estates, and audiences may object to the recreation of an actor’s face without consent.

Legal and intellectual property questions

Warner Bros. Discovery owns the rights to The Magnificent Ambersons. Any use of film footage, script elements, or character likenesses typically requires permission from the rights holder. Showrunner says it will not monetize the Ambersons project and that it would hand over materials to rights holders if there is a workable legal path, but its history of releasing unauthorized AI-made episodes makes the legal posture more complicated.

Key legal issues to watch

• Copyright ownership, including derivative works rights.
• Likeness and personality rights for deceased performers; rules vary by jurisdiction.
• Contractual rights tied to the original production, which may limit derivative uses.
• Platform policies where any reconstructed footage might be shared, which could block or label synthetic content.

Ethical concerns and cultural questions

Even if the legal issues are resolved, several ethical concerns remain. Recreating scenes using synthetic imagery prompts questions about authenticity, historical accuracy, and the respect owed to creators and performers who cannot give permission.

Main ethical points

• Authenticity, whether a reconstructed scene represents the original filmmaker’s intent or an approximation shaped by algorithmic choices.
• Consent for deceased artists, as their likenesses can be reproduced without their approval.
• Record keeping and transparency, so future viewers know how much of a restoration is original and how much is AI generated.
• Potential normalizing of deepfakes, which could make synthetic media more acceptable even when used for misrepresentation.

How historians and archives might respond

Museums, archives, and film scholars will likely push for rigorous documentation if AI-assisted reconstructions proceed. That means keeping source files, labeling reconstructed material clearly, and publishing the methodology used to create any new scenes. Academic standards for restoration typically emphasize faithfulness to original sources and full disclosure of speculative additions.

Industry context and broader trends

The Ambersons case fits a broader shift where startups and studios test AI tools for restoration, VFX, and creative production. Some companies are seeking deals with rights owners to legitimize their tools, while others have moved forward without authorization, prompting legal action or industry pushback.

What is changing

• More realistic face manipulation and scene synthesis tools, which can recreate visual styles and performances.
• Growing pressure on platforms and studios to set policies for synthetic reconstructions.
• Increased public discussion about whether and how AI should be used to maintain or reconstruct cultural heritage.

Key takeaways

• Showrunner plans to use a model called FILM-1, plus live actors and archival photos, to approximate lost footage from Orson Welles’ The Magnificent Ambersons.
• The project involves AI VFX artist Tom Clive and filmmaker Brian Rose, and Warner Bros. Discovery owns the film rights.
• Legal and ethical questions are central, including copyright, likeness rights, authenticity, and consent for deceased performers.
• Showrunner says it will not monetize the work and would seek a path with rights holders, but its prior unauthorized releases complicate trust.
• Any reconstruction will require clear documentation and labeling so audiences and scholars can tell what is original and what is AI made.

FAQ

Will this recreated footage be the actual lost material from Welles?

No. The plan is to create an approximation using AI and live-action to suggest how the missing scenes might have appeared. It will not restore the original cells of film that are gone, and it should be described as a reconstruction rather than an original scene.

Does Showrunner have permission from Warner Bros. Discovery?

Not publicly. Warner Bros. Discovery owns the rights. Showrunner says it would not monetize the project and that it would hand materials to rights holders if a legal path exists. Any public distribution without permission could raise legal challenges.

Are there rules about recreating an actor’s face after they have died?

That depends on law and local jurisdiction. Some places recognize posthumous rights of publicity or personality, while others do not. Families or estates can also object on ethical grounds. Consent from rights holders and estate representatives is important, even if not always legally required.

Concluding thoughts

The Showrunner plan to use FILM-1 to recreate lost scenes from The Magnificent Ambersons shows how generative AI can be applied to cultural heritage projects, and why such projects stir legal and ethical debates. The combination of AI keyframes, archival set photos, and face-manipulation techniques promises visually convincing output, but it also raises questions about who gets to decide what is restored and how reconstructions are labeled and shared.

For everyday viewers and film fans, the central takeaway is this. AI can expand the tools available for restoration, but transparent processes, rights holder engagement, and clear labeling will be essential to preserve trust and respect the history and people involved. The outcome will likely shape how future restorations use AI, and how the public expects reconstructed media to be presented.