AI and IP: The New Battleground
Author’s Note: While it is not the topic of this post, the ethical problems that plague AI usage are not lost on this author. To demonstrate, one paragraph from this post was created by ChatGPT after being prompted. The note at the end of the post will reveal which paragraph.
“One machine can do the work of fifty ordinary men. No machine can do the work of one extraordinary man.”
Whether American writer Elbert Hubbard birthed these words out of irrational fear or hard-learned wisdom, the rise of generative artificial intelligence (“AI”) has challenged this notion like no other innovation. It has done so by allowing machines to venture boldly into the realm once thought to be exclusively dominated by human: the arts. As modern philosophers work to understand what this progress means for society, this intrusion raises considerable legal questions about the practicability of intellectual property laws that were based on the formerly indisputable notion that only a living, breathing person is capable of creation.
Plainly, generative AI refers to algorithms and computer models that require “training” to work. This training is expensive and demands a tremendous amount of computational infrastructure, as AI relies on pools of inputted data to create. AI uses the data to, essentially, predict how a new piece of text, music, or visual scene should be composed and does so at incredible speeds.
It’s important to note that while AI-generated art can be highly creative and impressive, there’s ongoing debate about the nature of creativity and whether AI truly understands the art it generates. Some argue that AI lacks true consciousnesses, emotions, and intentions that human artists possess. Nonetheless, AI-generated art has gained recognition in art galleries, exhibitions, and even auctions, showing its impact on the art world and the broader creative landscape.
Creation is a process that generally has many sources of inspiration, where an author or artist pulls elements of various sources together and makes something new. As the old adage goes, good writers borrow; great writers steal. AI requires human beings to both train and prompt them to create anything. So is this not simply a more sophisticated paintbrush or pen? Arguably, it is, but the legal ramifications of creation go beyond the actual product.
Ownership of—and the ability to profit from—one’s work is the great carrot our Constitution envisions as being the driving motivation to compel creators to create. Already, AI is challenging conceptions about ownership. For instance, the Fourth Circuit recently dealt with a case that asked whether an AI could be an inventor for the purposes of obtaining a patent when an AI designer attempted to patent designs created by his personal AI, maintaining that he had did not contribute to the conception of the inventions and that any person could have taken the AI’s output and come up with the designs.
While the case was decided based on the statutory requirement that an inventor be a natural person, Thaler v. USPTO implied that AI must be limited to a collaborative role. Ownership, rather than creation, it seems, is the realm most reserved for mankind.
Even putting aside questions of ownership, there is a question of usage of other copyrighted works put into AI’s data pools. Since AI requires humans to input other works into the system for it to “learn” how to create, would this process be considered copyright infringement? Some would say yes, and in 2022, three artists filed a class action lawsuit in the Northern District of California against several generative AI platforms using this theory. Far from making a philosophical or public policy argument, the artists instead relied on market theory in their claim, arguing that the AI-created art negatively impacts their own market. Because some of the artists have distinct styles, the artists argue that AI platforms used their names for, essentially, marketing purposes, siphoning off income from the artists and diluting the value of their cultivated style by introducing similar-looking work. If the works created by the AI fail to sufficiently transform the existing, protected works, it is difficult to see how their use to make substantially similar works would not be considered copyright infringement. The lawsuit is ongoing.
Similarly, another lawsuit filed in 2023 focused on the use of AI platforms training with unlicensed works. For a platform to use unlicensed work and not be penalized for infringement, the work must be sufficiently derivative. Whether something is derivative enough depends on how a jurisdiction interprets the fair use doctrine, which, among other things, allows copyrighted works to be used without the owner’s permission—or license—if the work is transformative such that the copyrighted material is used in a way that was not originally contemplated by the creator.
The Supreme Court recently decided a potentially monumental case on fair use. The holding, which saw the Andy Warhol Foundation for the Visual Arts liable for copyright infringement, further outlines when a work is and is not sufficiently transformative enough to be afforded protection under the fair use doctrine.
Turning to AI, the question of fair use is one that will likely be decided in courts and legislative halls across the country, along with other pressing issues surrounding AI. Internationally, the European Union appears to be leading the charge, as landmark regulatory legislation makes its way toward passage.
As AI creeps into more and more aspects of life in this country, the arts will likely become a daunting battleground for proponents of AI and its detractors. In the face of this seemingly inevitable wave of litigation provoking challenges about our understanding of creation, intellectual property rights, and the market for artists, Hubbard’s words should have each of us asking one question above all else.
What is the worth of one extraordinary person’s work?
Author’s Note: The following paragraph used above was made by ChatGPT after being prompted:
“It’s important to note that while AI-generated art can be highly creative and impressive, there’s ongoing debate about the nature of creativity and whether AI truly understands the art it generates. Some argue that AI lacks true consciousnesses, emotions, and intentions that human artists possess. Nonetheless, AI-generated art has gained recognition in art galleries, exhibitions, and even auctions, showing its impact on the art world and the broader creative landscape.”
