Keywords: AI, art, media, copyright, creativity

The Baldy Center Blog Post 58
Blog Author: Lauren Wilson,  Asssociate Professor of Law, School of Law, University at Buffalo

Blog Post Title: New Technologies, Old Fears: AI, Copyright, and the Question of Artistic Value

"AI generated stuff generally looks ugly and disrespects talented actors, writer[s], artists, and people in general."
     –Anonymous, comment to the U.S. Copyright Office on      the copyrightability of AI-generated artwork (2023)

"I foresee a marked deterioration in American music and musical taste, an interruption in the musical development of the country, and a host of other injuries to music in its artistic manifestations."
     –John Philip Sousa, The Menace of Mechanical Music (1906)

The above two quotes, written more than a century apart, share a common throughline: art made with a particular technology should not be afforded copyright protection because the art made with the technology is aesthetically bad. When the United States Copyright Office requested public comments as it prepared its report on copyright and artificial intelligence in 2023, it received over 10,000 responses from creators, industry players, politicians, and other members of the public. [1]

The comment above was not an outlier. While many comments wrestled with whether AI-generated art could meet the requirements for copyright protection—“independently created” by a human author and having a “minimal degree of creativity” [2] —many others said things like “AI is ruining real artists. I am sick of seeing everything fake”; [3] or AI art will “literally take the soul out of everything. They’ll [sic] be no human connection and that is sad and scary.” [4] 

Humans have a long history of hating new technology. In a recent book, Techno Negative: A Long History of Refusing the Machine, Thomas DeKeyser writes that “an undercurrent of refusal . . . permeates the history of technology, at every step and turn." [5] When a new creative technology is introduced, questions quickly surface on whether art made with that technology is eligible for copyright protection. Conversations about these questions might begin with how the technology interacts with human authorship or other requirements for copyright protection, but they often get existential very fast. In his testimony before Congress on the copyrightability of sound recordings in 1906, John Philip Sousa—known primarily as a composer of patriotic marches like The Stars and Stripes Forever—stated that “these talking machines are going to ruin the artistic development of music in this country.” [6] Even more explicit, in a comment to the Copyright Office that opens by explaining that works created by AI should not receive copyright protection because AI authors are not human (a reasonable position, as far as copyright law is concerned), Eric Kristen closes by imploring the Office to “ensure humanity’s continued survival, because that is literally what’s at stake”. [7]

Although comments like this could be correct from an artistic point of view, they cannot be the reason that works created with new artistic technologies will or won’t receive copyright protection. Indeed, throughout the life of a copyrighted work, arbiters of copyright law expressly deny the law’s ability to make artistic judgments. At the registration stage, the Copyright Office states that it “will not consider the author’s skill, experience, or artistic judgment.” [8] Rather, the Office is “solely” concerned with “whether the work satisfies the originality requirement.” [9] In the courts, jurists have insisted for more than a century that it “would be a dangerous undertaking for persons trained only to the law to constitute themselves final judges of the worth of [insert any type of copyrightable art].” [10] Even in a case concerning a work that the presiding judge found offensive, the judge conceded that “[w]e are not called upon . . . to judge the quality of Defendant’s song, and whether we personally regard it as repulsive trash or a work of genius is immaterial to a determination of the legal issues presented to us.” [11] Yet this clearly established legal position does not prevent audiences from interpreting copyright decisions as judgments on the artists or works themselves. 

Even though comments like those I’ve mentioned here are not great arguments against copyright protection for technology-assisted works, they do offer a window into the public’s perceptions of the intersections between art, technology, and copyright. They reveal that, for many stakeholders, these intersections provoke major questions about the purpose of art, what it means to be creative, and even what it means to be human. Throughout history, new creative technologies have incited existential dread in creators and audiences of art, and the copyrightability of artwork created with new technologies seems to be an outlet for people to channel their fears about the changes the art that they hold dear may undergo. Former Register of Copyrights Barbara Ringer speculated in her lecture “The Demonology of Copyright” (incredible title) on whether there was “something special about copyright that provokes these strong and diametrically opposed expressions of feeling” so often voiced in debates about changes to copyright law. [12]

New technology draws us to feel big feelings, and AI is no exception. I think that both history and our present moment show that copyright law is often a place toward which to channel our anxieties about change. If we can just outlaw something, as many comments request of the Copyright Office, we won’t have to adapt to it, try to understand it, fight against it, or question the circumstances and corporations that have put us in this position of panic in the first place. There are certainly aspects of people’s concerns that copyright is well-positioned to handle. The Office can provide, for example, clarity on human authorship questions, or on the fair use of artists’ work in training data. It cannot, though, make commenters like those I quote here feel better about what art will look or sound like in 10 years or whether jobs in making incidental music for commercials or stock images for calendars will still exist. I hope they do. I also hope, though, that the conversations now focused on opposing copyright protection for AI-generated art on aesthetic or moral grounds might instead shift toward disciplines better positioned to handle them. When it comes to questions about our artistic and humanistic future, the humanities themselves might offer much more satisfying ideas.

ENDNOTES 

[1] U.S. Copyright Office, Notice: Artificial Intelligence and Copyright, Regulations.gov, here.

[2] Feist Publications, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 345 (1991).

[3] Stacie Stroup, Comment on Notice of Inquiry and Request for Comments on Artificial Intelligence and Copyright (Sep. 4, 2023), here.

[4] Charlene Chew, Comment on Notice of Inquiry and Request for Comments on Artificial Intelligence and Copyright (Aug. 31, 2023), here.

[5] Thomas Dekeyer, Techno-Negative: A Long History of Refusing the Machine 2 (2026).

[6] To Amend and Consolidate the Acts Respecting Copyright: Hearing on S. 6330 and H.R. 19853 Before the Committees on Patents, 59th Cong. 24 (1906) (statement of John Philip Sousa, composer; emphasis added).

[7] Eric Kristen, Comment on Notice of Inquiry and Request for Comments on Artificial Intelligence and Copyright (Aug. 30, 2023), here.

[8] U.S. Copyright Office, Compendium of U.S. Copyright Office Practices § 310.6 (3d ed. 2021).

[9] Id.

[10] Bleistein v. Donaldson Lithographing Co., 188 U.S. 239, 251 (1903).

[11] Parks v. LaFace Recs., 329 F.3d 437, 463 (6th Cir. 2003).

[12] Barbara A. Ringer, The Demonology of Copyright: Second of the R.R. Bowker Memorial Lectures 5 (1974).