Subscriber Benefit
As a subscriber you can listen to articles at work, in the car, or while you work out. Subscribe Now
Generative artificial intelligence systems, such as ChatGPT or OpenAI, have the potential to revolutionize the world as we know it. But copyright law may slam the brakes on the generative AI revolution.
Although generative AI is able to collect and process massive amounts of content to generate new expressions, these systems often do not align with the rights granted to owners of the original content used to create such new expressions.
Copyrights protect “original works of authorship fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device.” 17 U.S.C. § 102(a).
Such works of authorship include literary works, musical or dramatic works, pictorial, graphic or sculptural works, motion pictures and other audiovisual works, sound recordings, and architectural works.
The protections associated with a copyright are secured immediately upon creation of work of authorship. Fourth Estate Public Benefit Corp. v. Wall-Street.com, LLC, 139 S. Ct. 881, 887 (2019). “[A]uthorship is an entirely human endeavor. Authors of copyrightable works must be human.” Kelley v. Chi. Park Dist., 635 F.3d 290, 304 (7th Cir. 2011) (internal citations omitted).
The owner of a copyright has the exclusive rights either to perform or to authorize the following activities with respect to his or her copyrighted work: (1) reproduce the copyrighted work; (2) prepare derivative works based on the copyrighted work; (3) distribute copies of the copyrighted work to the public by sale or other methods of transfer; (4) in the case of literary, musical, dramatic, choreographic works, pantomimes, or motion pictures or other audiovisual works, perform copyrighted works publicly; (5) in the case of literary, musical, dramatic, choreographic works, pantomimes, or pictorial, graphic, or sculptural works, display the copyrighted work publicly; and (6) in the case of sound recordings, perform the copyrighted work publicly through digital audio transmission. 17 U.S.C. § 107.
The nature of generative AI conflicts significantly with these foundational copyright principles.
First, because generative AI by definition is not human, copyright rights and protections are generally not available to AI-generated works. The United States District Court for the District of Columbia recently held in Thaler v. Perlmutter that without the element of human creation, works generated by AI cannot be subject to copyright protections. Thaler v. Perlmutter, __ F. Supp. 3d __, 2023 U.S. Dist. LEXIS 145823, 2023 U.S.P.Q.2D (BNA) (D.D.C. Aug. 18, 2023).
While the court in Thaler recognized that copyright law is designed to adapt with the times, the court found that “[u]nderlying that adaptability … has been a consistent understanding that human creativity is the sine qua non at the core of copyrightability, even as that human creativity is channeled through new tools or into new media.” 2023 U.S. Dist. LEXIS 145823 at *10. As a result, such works are unlikely to receive the protections provided by copyright law.
Not only are AI-generated works not subject to copyright protections, but generative AI threatens to infringe on the rights provided to existing copyright holders.
Generative AI systems collect enormous amounts of data, which often includes copyrighted works, and apply that data to create new works. However, recent litigation could result in such applications constituting copyright infringement.
In Andy Warhol Foundation for the Visual Arts, Inc., v. Goldsmith, 143 S. Ct. 1258 (2023), the U.S. Supreme Court considered whether artist Andy Warhol’s colored silkscreens of photographer Lynn Goldsmith’s copyrighted black and white photograph of musician Prince constitutes infringement on Goldsmith’s copyright rights, or whether Warhol’s work constitutes a fair use of the photograph.
The Supreme Court in Goldsmith ultimately held that the purpose of Warhol’s colored silkscreen of the photograph generally had the same purpose as the original black and white photograph and therefore did not sufficiently transform the photograph to support a finding of fair use. Id. at 1264. As a result, the Supreme Court held that Warhol infringed on the copyright in the photograph. Id. at 1264-65.
Although Goldsmith does not explicitly relate to generative AI, the implications of this case could significantly impact the ability of generative AI systems to incorporate existing copyrighted material into works they create without the ability to rely on the doctrine of fair use as a defense.
The application of Goldsmith to AI will likely be tested quickly. On December 27, 2023, The New York Times Company initiated a federal lawsuit in the U.S. District Court for the Southern District of New York against OpenAI and Microsoft for copyright infringement based on their unauthorized use of The New York Times’ published works in their generative output.
The outcome of this case, especially in light of Goldsmith, may impact the ability of generative AI systems to use copyrighted works in any capacity, which could significantly hinder their ability to generate new works.
Generative AI presents limitless possibilities for creating works of expression, but an unknown future. However, while the future of generative AI remains unknown, one thing is clear—the future of AI will be shaped by its interaction with copyright law.•
__________
Jaclyn Flint is an associate at Riley Bennett Egloff LLP. Opinions expressed are those of the author.
Please enable JavaScript to view this content.