Artificial intelligence is moving deeper into Hollywood, but the industry’s biggest AI questions may ultimately have less to do with technology than with law. Studios, filmmakers, writers, actors, musicians and technology companies are experimenting with increasingly powerful AI systems capable of generating images, voices, scripts, video and other creative material. The possibilities are enormous, but so is the legal uncertainty. Who owns an AI-generated performance? Can a studio use an actor’s digital likeness after a contract ends? Can copyrighted material be used to train an AI model? What happens when an AI-generated character resembles an existing performer or creative work? And how should contracts change when technology can perform tasks that once required entire teams of human professionals? Hollywood is discovering that every technological shortcut can create a legal question behind it.
The industry has already learned that artificial intelligence cannot be treated as a distant possibility. During the 2023 Hollywood labor disputes, AI became a major issue in negotiations involving both writers and performers. The resulting agreements established important protections around the use of AI and digital replicas. Those negotiations were significant because they moved the discussion from theoretical concerns about future technology to concrete questions about employment, consent, compensation and control.
For performers, one of the most important issues is digital replication. Modern AI systems can potentially reproduce aspects of a person’s voice, appearance or performance. A studio may have legitimate reasons to use digital technology — for example, to complete visual-effects work, create a stunt double, alter an image or produce a character who appears younger. But a digital replica can also exist independently of the original performance. Once a sufficiently convincing digital version of a person exists, the legal question becomes much bigger: who controls it?
That question is particularly complicated because a person’s identity is not simply another piece of intellectual property. A face, voice and recognizable persona can have commercial value, but they are also connected to personal identity. Different jurisdictions treat publicity rights, privacy rights and related protections differently, creating a patchwork of rules rather than one universal standard. This means that the same AI-generated replica could potentially raise different legal questions depending on where it is created, distributed or used.
Contracts are therefore becoming one of Hollywood’s most important defenses against uncertainty.
Traditional entertainment agreements were designed around human performances and conventional production methods. AI introduces possibilities that may not have been contemplated when older contracts were written. A performer might agree to appear in a film, but does that agreement automatically authorize the studio to create a synthetic version of that performer for a future project? A writer might license a screenplay, but does that allow the material to be incorporated into an AI training dataset? A visual artist might create concept art for a production, but can that artwork subsequently be used to generate thousands of new images?
These are not merely philosophical questions. They can determine who receives money and who controls creative work.
The U.S. Copyright Office has been examining many of these issues through its ongoing initiative on copyright and artificial intelligence. Its work has included separate analysis of digital replicas, copyrightability and the use of copyrighted works in AI training. The agency’s examination reflects how quickly AI has created questions that existing copyright frameworks were not specifically designed to answer.
Copyright is particularly complicated because generative AI can produce something new while relying on enormous quantities of existing material. An AI system may generate a new image, paragraph, piece of music or video, but the system’s ability to do so can depend on patterns learned from existing works. That creates a fundamental legal debate over the relationship between training and infringement.
One side of the debate emphasizes the transformative potential of machine learning and the need for large datasets to develop useful AI systems. The other emphasizes the rights of creators whose work may be included in those datasets without direct permission or compensation. Courts, regulators and lawmakers are being asked to determine how existing legal principles should apply to a technology that can process information on a scale and speed that traditional creative industries have never encountered.
Hollywood has an especially large stake in the outcome because entertainment companies own enormous libraries of copyrighted material. Film studios, television networks, record companies and streaming services control valuable collections of scripts, performances, recordings, images and other creative works. If AI systems can learn from such material and generate new commercial content, the economic implications are substantial.
The question of ownership becomes even more complicated once an AI system actually produces something.
Copyright law generally requires human authorship for copyright protection in the United States. The U.S. Copyright Office has stated that material generated entirely by AI is not eligible for copyright protection merely because someone prompted the system. However, human contributions can potentially be protected when a person contributes sufficient creative expression or arranges AI-generated material in an original way.
That creates an unusual situation for entertainment companies. A studio could spend substantial money creating an AI-assisted film sequence, but the legal protection available for different components of that sequence may depend on how much human creative contribution was involved. The more AI becomes involved in production, the more important it may become to document the human decisions behind a finished work.
Record keeping could therefore become a surprisingly important part of future entertainment production.
Studios may need to document which tools were used, what material was supplied to an AI system, who created prompts, who edited the resulting output and which human decisions shaped the final product. That information could become valuable if ownership is later challenged. What once looked like a technical production detail could become evidence in a copyright dispute.
There is another major issue: unauthorized digital replicas.
Imagine an actor discovers that a synthetic version of their face or voice has appeared in a commercial, advertisement or entertainment project without permission. The actor may have contractual claims, publicity-rights claims or other potential legal remedies depending on the circumstances. But technology makes enforcement more complicated because digital replicas can be created and distributed quickly.
The legal system is beginning to respond.
In the United States, lawmakers have discussed federal legislation addressing unauthorized digital replicas, while individual states have also adopted or considered laws concerning digitally replicated voices and likenesses. California, for example, has enacted legislation addressing certain uses of digital replicas, reflecting growing concern about the ability of AI to imitate individuals without authorization.
The challenge is creating rules that protect people without preventing legitimate creative expression.
A filmmaker may want to digitally recreate a historical figure for a documentary or fictional story. An actor may want to license a digital likeness for a game. A musician may want to authorize an AI-generated version of their voice for a specific project. These uses are not inherently harmful. In fact, some could create new creative opportunities and revenue streams.
The legal question is therefore not simply whether digital replicas should exist.
It is who has the authority to create them, under what conditions and for what purpose.
That distinction will likely shape the next generation of entertainment contracts.
Consent may become more detailed. Contracts could specify exactly how a person’s likeness can be reproduced, which media can use it, how long authorization lasts and whether the digital representation can be modified. Compensation structures could also evolve. Instead of paying only for a performance, companies may negotiate separate payments for digital-replica rights.
Writers face a similar transformation.
AI can assist with brainstorming, research, drafting and rewriting, but writers’ organizations have pushed for protections against studios using AI to undermine human writers or exploit their work. The Writers Guild of America agreement reached after the 2023 strike established provisions governing AI use, including rules concerning the use of writers’ material and AI-generated writing.
The significance extends beyond screenwriters. It raises a broader question about what it means to be the author of entertainment in an age of machine-assisted creation.
If a writer develops a story, uses an AI tool to generate alternative dialogue, rewrites that material and then produces the final screenplay, the human creative contribution remains substantial. But if a company simply asks a model to generate an entire screenplay based on a prompt, the legal and creative status of the resulting work becomes more complicated.
Studios may also have to consider disclosure.
Audiences increasingly care about how entertainment is made. Some viewers may have no problem with AI-assisted visual effects or localization. Others may object to synthetic performances or AI-generated creative material. Companies may therefore find themselves under pressure to explain when AI has been used, particularly when the technology directly affects a performer’s identity or the authorship of a work.
Transparency could eventually become part of the entertainment industry’s trust equation.
There is also a business risk that is easy to overlook: litigation can erase the financial advantage AI promises to create.
A studio might adopt an AI system because it appears cheaper and faster than traditional production. But if the resulting work triggers a major copyright lawsuit, likeness dispute or contractual conflict, the cost savings can disappear quickly. Legal uncertainty can become a business expense of its own.
This is why companies are increasingly likely to establish internal AI policies.
Those policies may govern which tools employees can use, what information can be entered into AI systems, whether copyrighted material can be uploaded, how generated content must be documented and who has authority to approve AI-assisted work. Large entertainment companies may have legal and compliance teams specifically focused on AI-related production risks.
Smaller filmmakers face a different challenge.
Independent creators may have fewer legal resources but increasingly have access to powerful AI tools. That democratization can be exciting because it allows small teams to create sophisticated content. However, independent creators may also be more vulnerable if they unknowingly use copyrighted material, replicate someone’s likeness or violate a platform’s terms.
Education will therefore become increasingly important.
Filmmakers need to understand not just what AI tools can do but what rights they are using in the process. A tool may be technically capable of generating a celebrity likeness, for example, but technical capability does not automatically mean legal permission. The same principle applies to copyrighted characters, music, photographs and other recognizable creative properties.
The law is also likely to evolve unevenly.
Federal legislation, state laws, court decisions, agency guidance and industry contracts may all address different parts of the problem. That could create an environment where the rules change rapidly and differ across jurisdictions. Entertainment companies operating globally will have to navigate multiple legal systems simultaneously.
Europe, the United States and other regions may develop different approaches to AI regulation, privacy and copyright. A production distributed internationally could therefore face different obligations depending on where its content is created and where it is made available.
This could eventually influence where entertainment companies choose to conduct certain AI-related work.
Legal predictability has always been important to Hollywood. Productions involve enormous budgets, complicated intellectual-property arrangements and international distribution. Companies need to know who owns what before they invest millions of dollars in a project. AI introduces another layer of uncertainty into an already complicated system.
But there is an opportunity hidden inside the legal uncertainty.
Clearer rules could make responsible AI adoption easier.
If performers know that their likeness cannot be replicated without consent, writers know how their work can be used, artists understand how their material is protected and studios know what documentation they need to maintain, companies may become more confident about using AI responsibly. Legal safeguards do not necessarily have to prevent innovation. They can establish boundaries that allow innovation to happen with greater certainty.
The future of Hollywood may therefore depend on a new balance.
The industry needs enough flexibility to experiment with emerging technology while maintaining meaningful protections for the people whose creative work makes entertainment valuable. AI can accelerate production, reduce certain costs and create new possibilities. But speed cannot become a substitute for consent, and efficiency cannot automatically override ownership.
The most important legal battles may ultimately happen before a movie reaches an audience.
They may happen when contracts are negotiated, datasets are assembled, models are trained, digital replicas are authorized and production decisions are documented. By the time a finished AI-assisted film appears on a streaming platform, many of the most important legal questions may already have been decided behind the scenes.
That is what makes Hollywood’s AI future such a complicated legal minefield.
The technology itself is advancing rapidly, but the rules governing its use are still catching up. Every new capability creates another question about rights, responsibility and control.
Can AI help make a movie?
Absolutely.
Can it transform production?
Already happening.
Can companies use it however they want?
That is where the legal battle begins.
Hollywood’s AI future will not be determined by technology alone. It will be shaped by contracts, courtrooms, regulators, unions, lawmakers and the creators whose work gives the entertainment industry its value.
The machines may be learning how to create.
The law is now learning how to respond.
Caption: Hollywood’s AI revolution is moving fast — but the legal questions may move even faster.
References
U.S. Copyright Office — Copyright and Artificial Intelligence
U.S. Copyright Office — AI Initiative Reports
SAG-AFTRA — Artificial Intelligence Resources
Writers Guild of America — 2023 MBA Artificial Intelligence Provisions
Leave a Reply