Photo by Andrea De Santis on Unsplash
Introduction
As AI-generated content becomes more prevalent, there is a growing need for a legal framework that ensures creators, developers, and innovators can benefit from their contributions while encouraging further advancements in technology. The rise of AI offers the possibility of expanding human creativity, enabling new forms of expression and collaboration. However, this expansion also calls for clarity on ownership, licensing, and the protection of human-created content. It is crucial that we establish principles that recognize both the value of human creativity and the role of AI as a tool for enhancing it.
Generative AI is revolutionizing the world of art by producing vivid and complex works that have already found its way into exhibits. For example, The Museum of Modern Art in New York and Mauritshuis in the Hague have both displayed AI-generated work in its galleries. AI tools such as Stable Diffusion, Midjourney, and DALL·E 2 can create incredibly realistic pieces of art with vast datasets. Researchers found that people generally cannot tell the difference between AI-generated art and human art. In 2022, an AI-generated work won first place in the “digital arts/digitally-manipulated photography” category at the Colorado State Fair Fine Arts Competition. However, this accomplishment for AI-generated art is controversial because critics see it as unfair competition. Moreover, as this new art medium evolves, complex legal issues also arise. Questions surrounding ownership and whether AI-generated art constitutes “derivative works” are the most prevalent. As of now, AI-generated content cannot be copyrighted. A central issue is if there were to be copyright protection for this material, would it apply to the input data or the final result? As we further explore AI-generated art and its legal implications, safeguarding the delicate balance between intellectual property rights and artistic creativity will be crucial.
The AI-generated artwork entered by Jason Allen into the Colorado State Fair Image: Jason Allen via Discord
Defining Human Authorship
Litigation in AI-generated content is quickly on the rise. In 2022, artists sued AI platforms, alleging that their works were used to train AI without their permission. Therefore, the product was an unauthorized “derivative work.” For infringement purposes, a derivative work must copy an original work in some way. If the court finds that these works are not significantly different from the original works, they could be deemed infringing. The Supreme Court has focused on the term “transformative use” for when a work is not “derivative." If a work is found to be a transformative use, then it is copyrightable and safe from claims of infringement. In Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith, the core issue was whether the use of a photo was “transformative” under the doctrine of fair use, meaning whether it was different enough from the original work to be its own independent work of authorship. The Court found that although the photograph was artistically altered, it was not sufficiently transformative to be exempt from copyright infringement because it was closely tied to the original photo, and therefore, it constitutes a derivative work.
AI-generated works may be at risk of being classified as derivative works like in the Andy Warhol case. AI-generated art is already being criticized as stealing from other artists and copying their styles. However, the end result of AI-generated art is its own creation, merely relying on others’ work for inspiration—or input data. Still, copyrights are not being afforded to creative works that do not have human authors—likely out of fear that granting those copyrights will lead to increased litigation.
AI-Generated Art in Classic Art Museums
Supporting AI-generated art in classic museums while honoring its unique value, despite its current lack of copyrightability, presents both challenges and opportunities. However, there are ways in which we can integrate and recognize AI-generated art in museum spaces while also respecting its differences from traditionally copyrightable work. The Museum of Modern Art (MOMA) hosted an AI-generated installation from the museum’s own collection, and the Mauritshuis in The Hague hung an AI variant of Vermeer’s Girl with a Pearl Earring while the original was away on loan. The discussion around these exhibitions touches on how AI art challenges traditional ideas of creativity and authorship. Even though AI-generated works are not eligible for copyright protection under current U.S. law, we can still honor these creations by treating them as collaborative works between humans and machines. Exhibiting such art offering curatorial transparency and fostering public discourse about the role of AI in art are key ways to honor and validate AI’s role in creative expression.
Curatorial transparency means communicating to audiences how AI-generated art is created and its relationship with the original works it draws from. For example, at the MOMA, the AI installation was built from the museum’s existing collection, making the relationship between traditional and AI-generated works clear to the viewers. Similarly, the Mauritshuis showcased an AI-generated version of Girl with a Pearl Earring, ensuring visitors understood the distinction between the original masterpiece and the AI variant. Transparency not only helps visitors appreciate the technological and artistic process involved but also builds trust by openly acknowledging the differences between human and AI creation. Public discourse around these exhibitions helps museums and viewers think about the broader legal and ethical questions that AI-generated art poses. Exhibits like those at the MOMA and the Mauritshuis encourage the audience to question and discuss what constitutes “art” in the digital age. The controversy surrounding the AI version of Girl with a Pearl Earring shows how such displays can engage the public, inviting diverse opinions and reflections on the relationship between AI and traditional art. The interaction with the public feeds into larger conversations about the limits of copyright law in protecting new forms of digital expression.
As AI continues to revolutionize art, the legal framework must adapt, but in the meantime, institutions can honor AI-generated art through transparency, education, and engagement. Exhibitions that spark discourse around these themes can play a pivotal role in how society navigates the AI revolution and its impact on the traditional boundaries of artistic and intellectual property.
The Copyright Office’s Stance on AI-Generated Works
The U.S. Copyright Office’s 2023 initiative addressed a pivotal issue in modern copyright law: the role of artificial intelligence in the creation of works and the use of copyrighted materials in AI training. The Copyrights Office’s stance and current policy underscore the need for thoughtful consideration of these issues, balancing the rights of human creators and the growing influence of AI in creative fields. Historically, U.S. Copyright Law has been built on the premise that works must be the product of human creativity. The U.S. Copyright Act defines “works of authorship” as creations that originate from human intellect, including literary, musical, dramatic, and visual arts.
While AI systems have developed to the point of autonomously generating music, art, literature, and other forms of creative content, the boundary between human and machine-generated creativity has blurred. In response to this technological evolution, the Copyright Office has reaffirmed its long-standing principle that copyright protection applies only to works with a sufficient degree of human authorship. When AI “determines the expressive element of its output, the generated material is not the product of human authorship.” The Copyright Office holds its position that, while AI can be used as a tool by human creators, works fully generated by AI without human involvement are not eligible for copyright protection. This stance places significant limitations on the scope of copyright for AI-generated content. To qualify for copyright protection, creators utilizing AI tools must demonstrate meaningful human involvement in the creation process. This requires the human creator to make decisions about key creative elements. However, establishing this level of creative decision-making can be complex. The U.S. Copyright Office explained that AI generative tools, such as Midjourney, are tools that can be guided by an author to reach their desired image “because it generates images in an unpredictable way.” As a result, obtaining copyright protection for works created with AI requires clear evidence of human authorship and creative input, despite the challenges posed by the unpredictable nature of AI-generated content.
Policy Considerations and the Road Ahead
The Copyright Office’s 2023 Initiative underscores the need for policymakers to carefully examine the intersection of AI technologies and copyright law. While the Office has maintained a clear stance on the requirement for human authorship in copyrightable works, it recognizes that current regulations may not fully address the complexities introduced by AI.
One potential solution could involve legislative reforms to clarify the scope of copyrighted AI-generated works and the use of copyrighted materials in AI training. Some have proposed that new categories of copyright rights could be created to address the unique nature of AI-generated content. Others suggest that licensing frameworks could be developed to allow for the lawful use of copyrighted materials in training datasets, ensuring that creators are compensated for the use of their works. The U.S. Copyright Office has also called for public input as part of its initiative, recognizing the need for diverse perspectives from creators, technology companies, legal experts, and the public at large. This collaborative approach reflects the Office’s commitment to crafting policies that balance the rights of human creators with the growing influence of AI in creative industries.
The Copyright Office is positioned to lead the way in crafting guidelines that address these emerging challenges. By considering the complexities of AI-generated works and engaging with a wide range of stakeholders, the Copyright Office can foster a system that balances the rights of creators with the evolving nature of content creation. This decision will not only protect artistic and intellectual property but also promote innovation in ways that respect the contributions of all involved parties.
*The views expressed in this article do not represent the views of Santa Clara University.
Comments