Who Owns Ideas? Humans versus AI in Intellectual Property
The Artificial intelligence actress Tilly Norwood was created in February of 2025 by the UK-based production company Particle6. She looks and acts completely human and, since her creation, has been in a few short films with a large presence on social media. However, since her first appearance, the AI-actress has faced severe backlash across social media. Norwood’s debut was only the beginning of a sweeping campaign with AI emerging in the entertainment industry. As NBC points out, AI filmmaking over the past few years has quickly entered mainstream media. Studios have begun working with AI companies to explore the technology’s potential in content creation, and apps like Sora by ChatGPT allow people to create, upload, and view AI-generated videos featuring anyone via a face-scan and text prompt. The use of AI across art media has quickly become a polarizing topic for many artists. In an age where AI has inevitably found its way into entertainment and art industries, how does intellectual property law create boundaries that protect human-generated art, and where, if at all, is the boundary drawn on what can be protected as intellectual property when AI is used? With AI rapidly evolving, intellectual property law must be able to distinguish between artists who use AI and works of art that are simply machine-generated, in order to protect human creativity and ingenuity without blocking the use of technology.
Historically, copyright law has been based on the premise that creative works are produced by humans, but as AI has become an increasingly popular tool for artists, this idea has come under strain. In Title 17 Chapter 1 Section 102 of U.S. copyright law, it is stated that in order for something to be protected by copyright, it must possess “original works of authorship fixed in any tangible medium of expression”. In this law, “authorship” means that the work must be human-made, with a clear level of creativity. In the case of AI, it can legally learn ideas, styles, and other information from copyright-protected material, as long as it is not copying intellectual property. But, because AI is not human, AI lacks this definition of “authorship,” and therefore any work created by AI does not qualify for copyright protection. The reasoning behind this, as explained in a 2023 congressional hearing, was to ensure that human creativity is preserved and rewarded. However, as AI becomes a common tool that pushes creativity further, what happens to works created by humans with the help of AI?
Recent lawsuits and court cases from the past years can shed light on the issue of blurred IP laws in response to AI. One example that illustrates this issue is the ongoing Midjourney Lawsuit, headed by Walt Disney and Co. and Universal Pictures, for using characters originally made and patented by the companies. The entertainment companies accuse Midjourney of “blatantly” profiting off of characters from the major franchises each company produced, including “Star Wars”, “Marvel”, and “Cars”. This lawsuit not only reveals how serious a threat AI poses to companies that have built their brand around characters and stories they created, but, more deeply, shows the importance of preserving and respecting human creativity in media.
The issue of human creativity becomes even clearer/more evident in the case of the comic book “Zarya of the Dawn,” which was rejected for copyright protection because the author used AI to create images in the book. In a letter, the copyright office reasoned that copyright law protects “the fruits of intellectual labor” that “are founded in the creative powers of the mind”. According to the copyright office, Zarya of the Dawn’s illustrations were not “original intellectual conceptions of the author”. This creates the precedent that works of art must stem from a high level of human and intellectual creativity.
However, according to the case of Théâtre D’opéra Spatial, works must have more than just human creativity. This AI artwork was created with Midjourney and denied for copyright registration numerous times because of a lack of human involvement. The co-creator claimed that he worked with the AI by giving it prompts and adjusting those prompts according to what the AI produced. Regardless, the Copyright Review Board said the author had “no control” over how AI responded to the prompts given, and therefore could not be registered. However, the co-creator continuously had the AI alter the image to get what he envisioned, which adheres to the precedent created from “Zarya of the Dawn”. From this, another interpretation is added to what can be protected, with this one requiring physical work by the author.
Together, this raises the question of where the line should be drawn on AI involvement. The website for copyright registration explains this by asking if the “‘work’ is… one of human authorship, with the computer [or other device] merely being an assisting instrument” or, if the “elements of authorship in the work… were actually conceived and executed not by man but by a machine.” According to the U.S. Copyright Office, to meet the requirements for copyright protection, the work must be conceptually and physically created by the author, rather than mostly produced by AI and edited by the author. Still, the different interpretations of what qualifies for copyright protection show how ambiguous the line is for AI in relation to IP, suggesting that perhaps intellectual property law should be moving in a new direction to become clearer and accommodate the growing use of AI.
Because AI has complicated the question of who creates ideas, laws surrounding intellectual property should shift to consider the amount of human input and control when creating art. On this topic, many ideas have come to light in an attempt to solve the issue of human authorship amidst the era of AI. One proposal comes from a University of Florida Law Review article by Edward Lee. Lee’s problem is that current copyright laws do not incentivize creating new works with AI “lest they jeopardize their eligibility for copyrights” (137). The solution Lee has is for artists to receive copyright protection of AI works as long as they contain a “minimally creative selection or arrangement of elements” from the author (137). Requiring a bare minimum will push authors to go through “numerous iterations of works… making refinements and a selection or arrangement of the elements”. In doing this, the author will likely meet the bare minimum needed while still pushing creativity with the use of technology. If this idea works as Lee hopes, it could be immensely helpful in creating an environment that allows artists to push the boundaries of art and entertainment without surrendering artistic integrity to AI. Similarly, an article by the New York State Bar Association (NYSBA) has suggested changes to patent law to allow the use of AI as a technological tool, like any other technological tool that has significantly evolved over the past decade. It suggests laws like the “Elvis Act,” which forbids the replication of a person’s voice or qualities through AI, as well as establishing clearer standards for what constitutes “training” AI, and overall reforms intended to preserve human ingenuity as the most important part of copyright law, while allowing people to use AI to push creativity further. This idea is a more involved approach as it will actively ban parts of AI, allowing for clearer rules than a “bare-minimum requirement”. However, both of these articles give valid approaches to protect human ingenuity while still encouraging technological innovation that helps push creativity.
As AI continues to grow and reshape how art is created, it has exposed massive gaps in intellectual property laws. Disputes for copyright protection, such as Zarya of the Dawn and Théâtre D’opéra Spatial, reveal how different definitions of human authorship arise based on U.S. Copyright Law. Lee and the NYSBA have suggested reforms to clear these grey legal areas, such as a bare-minimum requirement and laws that safeguard pre-existing artists’ and characters’ features. Although these may not be complete solutions, they nevertheless give way to possibilities that allow intellectual property laws to keep up with the rapid expansion of AI. As society and AI move forward, one thing becomes clear: Intellectual property laws must evolve to allow for creativity with the use of AI.