Translation

Generative AI and Copyright Laws: Bridging the Divide

Generative AI and Copyright Laws: Bridging the Divide

Siyao Cheng & Howard Graves

AI in our everyday lives

In recent years, “artificial intelligence” (AI) has been the nucleus of myriad conversations, including discussions on using information, sparking particular concerns within the realm of copyright. There is a vast array of AI technologies developed by various companies, each specializing in performing tasks assigned by human beings. Generative AI, a class of AI technologies like ChatGPT, produces images, texts, audio, and video from large databases when we give it instructions. It is also poised to take on significant roles in industries. For example, The New York Times plans to utilize Generative AI to help with news reporting. Overall, it has been increasingly integrated into almost all aspects of our lives, offering convenience and efficiency, as reported by the Pew Research Center

Nevertheless, it is associated with a flip side. One of the concerns is that Generative AI’s output in response to user prompts often draws from copyrighted works, eliciting a question nowadays: Can the content created by Generative AI qualify for copyright protection itself? Here, we strive to extend our insights into the question.

—Can the content created by Generative AI qualify for copyright protection itself?—

AI in the copyright regime

Copyright laws are typically the first method that comes to mind when dealing with the use and distribution of information. However, AI’s produced content does not straightforwardly reconcile with the copyright requirements due to its unique nature and the way it operates. Copyright laws typically protect works that are fixed in a tangible medium, display a minimal level of originality and creativity, and are authored by humans.

On one hand, Generative AI-produced content, triggered by user prompts, manifests in a digital, fixed format that satisfies the criterion of fixation. On the other hand, the fundamental purpose of copyright is to encourage ongoing creativity and innovation. Most later copyrighted works build upon earlier copyrighted works. AI seems to work in a way that generates content by gathering and combining existing, diverse sources from large databases, but all it does is just pull out all of the relevant sources together for users. The content generated by AI does not always meet the threshold of creativity and originality requirement for copyright protection. Additionally, copyright laws require human authorship—meaning they grant protection to works originating from the human creators. AI-generated content, however, is created on the basis of algorithms or machines and lacks a human author in the traditional sense. 

These incompatibilities compel us to further explore the nexus between AI-generated content and copyright laws. We think of whether copyright should be the law that plays a vital role in the protection of AI-generated content. If so, how might copyright laws adapt to accommodate the unique capabilities and outputs of Generative AI, especially considering its widespread influence on people’s daily lives? 

Exploring the possibility of copyright protection for nonhuman entities AI

In 1735, Benjamin Franklin was writing about fire prevention for the townspeople of Philadelphia when he wrote a statement that would become a well-known maxim: “An ounce of prevention is worth a pound of cure.” That piece of wisdom could now be applied to the likelihood of paradigm-shattering disputes involving AI, the U.S. Copyright Office, Congress, and the public. In 2023, the federal district court of the District of Columbia stated that only human beings are copyright-eligible for intellectual property (Singh). This has always been the accepted, common sense view; it is the paradigm. The statement, as is, almost goes without saying. 

Paradigms are powerful. They are pervasive and usually exist without a general awareness of their existence, but eventually, paradigms grow old and tired and lose their energy. Such a situation could be occurring today with the growth of AI. Can AI create copyrightable, unique literary works, architectural designs, or musical compositions? According to federal interpretation, it cannot because AI at the present time is dependent on human input and programming. However, copyright laws have been revised numerous times during the past century and a half and will continue to be revised. Furthermore, a common dictionary definition of “creation” is the act and process of bringing something into existence. No definition has ever existed that specifies creation being dependent on human involvement—until the advent of AI. Copyright laws have had to revise the definition of “creation” in order to adapt to AI technology. AI is challenging the paradigm of what creation is, or at least how we should think about it. Admittedly, AI is in its infancy and heavily dependent on human input and oversight, but because of the nature of its purpose, AI could potentially evolve a unique version of intelligence with its own communication methods, artistic techniques, and viewpoints about its existence. After all, “Intelligence” is the second term in its name, and the first term, “Artificial,” refers to a computer program that can do things that previously only humans could do. In other words, AI can do things that humans also do. 

It can be argued that technology has become a fifth social institution along with education, family, religion, and government. Each of these provides a unique model for thought, behavior and general rules by which to engage with societal life. It could also be argued that computers and their associated technologies have become non–human members of society because of our daily interactions with, and dependence on, their presence. AI is a burgeoning field punctuated by outspoken people who are holding their arms up and strenuously waving to all of us with significant messages. These are not necessarily dire warnings but rather cautions about the need to prepare for a new kind of cognitive cohabitation involving humans and AI. 

Confronting the Future (Which Has Already Arrived)

Generative AI programs are the possible harbingers of a paradigm shift concerning the nature of intellectual creation and copyrightable intellectual property. These early AI programs are baby steps, but baby steps toward what? We don’t know, but just as Ben Franklin admonished the people of Philadelphia to look ahead and think about how to prevent destructive fires, our social institutions should look ahead and think about how to assimilate AI into a matrix of inclusivity. We should assess the copyright landscape for potential AI “fires” and do what we can to eliminate future legislative and legal fires.

Cite this article in APA as: Cheng, S. & Graves, H. Generative AI and copyright laws: Bridging the divide. (2024, July 9). Information Matters, Vol. 4, Issue 7. https://informationmatters.org/2024/07/generative-ai-and-copyright-laws-bridging-the-divide/

Author

  • Siyao Cheng

    Siyao Cheng is a doctoral student at the School of Information Sciences at the University of Illinois at Urbana-Champaign. Her research focus is everyday information behavior and copyright education.

    View all posts

Siyao Cheng

Siyao Cheng is a doctoral student at the School of Information Sciences at the University of Illinois at Urbana-Champaign. Her research focus is everyday information behavior and copyright education.

Leave a Reply