Imagine opening Netflix in 2035 and finding no catalog of movies to choose from. Instead, you type: “Make me a two-hour science-fiction thriller with the pacing of a Christopher Nolan film, the humor of a Marvel movie, and an ending where the protagonist survives.” Minutes later, your movie is ready. You can revise the plot, change the ending, or ask for a sequel starring a different character.
As Generative Artificial Intelligence (Gen AI) continues to improve at creating content in the mediums we use to consume entertainment material, and as we get better at using that Gen AI, this is a very possible—and even probable—future. Beyond a personalized film, this technology could produce personalized novels, video games, songs, and more. Rather than the consumer choosing among a finite collection of completed works based on individual preferences, those preferences could be used to generate endless entertainment tailored entirely to them.
This possibility raises a fundamental question for copyright law and economics of entertainment: If entertainment becomes infinitely customizable, what exactly is the “work” that copyright is supposed to protect?
The traditional entertainment industry revolves around fixed works. Studios finance movies, publishers release books, and developers create games, expecting to recoup their investments by distributing those works to large audiences. Copyright law supports this model by granting creators exclusive rights to reproduce, distribute, and adapt the expression of their work.
Generative AI could disrupt that arrangement. If a high-quality movie can be generated on demand at relatively little additional cost, the finished movie may become less economically valuable than the creative assets behind it. Characters, fictional universes, distinctive settings, and established franchises could become the foundations for countless personalized experiences.
This would not mean that copyright suddenly begins protecting creative building blocks. Copyright already protects certain recurring elements, including sufficiently distinctive fictional characters. In DC Comics v. Towle, the Ninth Circuit recognized copyright protection for the Batmobile as a sufficiently delineated character. At the same time, copyright generally does not protect abstract ideas, genres, or artistic styles. As the Second Circuit explained in Nichols v. Universal Pictures Corp., distinguishing protectable expression from increasingly abstract ideas is a central challenge of copyright law.
The question, then, is whether personalized AI entertainment will make those ingredients more economically important than the finished works in which they appear. Will the source of value derived from copyright protection fundamentally shift and, if so, take the law with it?
An important underlying fact is that Gen AI outputs are protectable only to the extent that a human determines sufficient expressive elements. Merely providing prompts, under current technology, generally does not establish the human authorship necessary for copyright protection. Thus, when creating a work, the AI provider does not automatically become the author simply because it built it, and the user does not automatically become the author simply because they requested it.
That result could create a striking incongruity: consumers might enjoy increasingly sophisticated entertainment without owning copyright in the works they generate. Thus, one consequence might be that platforms begin to rely even more on subscriptions, contracts, or technical restrictions to control access and use. Where does copyright come in the mix? If existing intellectual property is incorporated in this process, an absence of copyright in a new output does not make the underlying copyrighted material free to use.
Copyright’s traditional economic rationale is to encourage the creation and dissemination of works by granting creators limited exclusive rights. Based on the underlying notions of excludability and nonrivalrousness, this, so far, has meant a focus on ownership of final works, because that is what consumers actually want. But when generating another final work becomes cheap, while the underlying creative assets remain protected and scarce, the value of entertainment would migrate toward the ingredients that make generation desirable: recognizable characters, fictional worlds, distinct props, and maybe even the very systems capable of producing compelling experiences.
As licensing demand shifts along with this pattern, rights-holders will follow the compensation, and we may exist in a world tailored to the system of tailoring. That is, right-holders will shift focus to solely produce catalogs of these ingredients for use by Gen AI systems, implying the chilling of the comprehensive creative work.
There is already scholarship exploring personalization and copyright from another direction. In Toward the Personalization of Copyright Law, Adi Libson and Gideon Parchomovsky examine whether copyright rules and remedies should be tailored to individual users. Their proposal concerns personalizing the law, rather than personalizing the works themselves. But it offers a useful conceptual parallel: copyright’s relationship with individual consumers need not remain fixed.
The likely legal response is not necessarily to abandon the concept of copyrighted work. Under 17 U.S.C. § 102, copyright protects original expression rather than ideas, procedures, or systems; 17 U.S.C. § 106 establishes the exclusive rights that copyright owners hold, including rights involving reproduction and derivative works. These existing categories may remain central, albeit malleable, even as the economics of entertainment changes. The deeper question informing the extent of this flexibility is whether copyright will continue to organize entertainment primarily around identifiable works when technology makes those works fluid, personal, and endlessly reproducible. In an age of bespoke entertainment, the most valuable intellectual property may no longer be the blockbuster itself, but the creative universe from which every viewer can make their own.
None of this means this economic outcome is inevitable or copyright must be fundamentally overhauled. Audiences may continue to value human performances, trust in curated releases, and participate in shared cultural experiences. Studios could retain their influence through exclusive content, proprietary platforms, or control over access to popular fictional universes. AI might well stay its course in the transformation of the entertainment industry without eliminating its traditional players. But in a world largely defined by uncertainty surrounding the future of artificial intelligence, it is worth questioning whether the industry will cease functioning as we know it.