Andy Warhol in 1981. Photo: Nora Schuster/Imagno/Getty Images.
WAS IT REASONABLY NECESSARY? This is a question asked every day in law schools and courthouses. Less in MFA programs, artist studios and museum acquisition committee meetings. Imagine a lawyer standing right behind Andy Warhol in The Factory, asking, “Does it have to be Brillo?” What if you had just created your own soap box? Or imagine MoMA turning down a trove of image-generating works because the law has questions, not about provenance, but about copyright clearance. We could be moving in that direction, now that the Supreme Court has introduced a standard of reasonable necessity into transformative fair dealing analyzes with its recent decision In Andy Warhol Foundation for the Visual Arts vs. Goldsmith.
The case grew out of an esoteric pattern of fact – what in copyright circles is considered use beyond the scope of a license. In 1984, Warhol created a serigraph of Prince based on a source photograph taken in 1981 by Lynn Goldsmith, with her express permission. He did not, however, have permission to make the additional fifteen works (mainly serigraphs and some pencil drawings) that he also created from the same image. When Goldsmith later learned that the Warhol Foundation had allowed Condé Nast to use one of these additional works, known as orange prince, 1984, she alleged copyright infringement and, as alleged by Warhol’s attorney during oral argument, demanded seven figures and the transfer of Warhol’s copyright. The Foundation, not lacking in nerve, responded by suing Goldsmith in federal court and asking the court to declare that Warhol’s unlicensed uses were fair under the Copyright Act.
In a long, spirited opinion filled with bizarre bedfellows, accusatory footnotes, dissenting art history lessons and photoshopped imagery, the Supreme Court recently upheld a ruling in Goldsmith’s favour. . With the notable exception of legal scholar Amy Adler’s answer to the decision in Art in America, reactions in both popular media and academic circles are best described as uncertain. The title of Blake Gopnik’s play in the New York Times summarizes this ambiguous post-decision atmosphere: “A decision against Warhol should not hurt the artists. But it could.
This ambivalence results in large part from the way in which the majority opinion has it both ways. Obviously, Warhol’s fair use defense was dismissed and Goldsmith prevailed. It’s a clear victory for photographers and a defeat for artists who work with copyrighted sources. If you have any doubts about this conclusion, look at how the nonpartisan amicus briefs aligned themselves with the respective camps before the decision (American Society of Media Photographers for Goldsmith; Robert Rauschenberg, Roy Lichtenstein and Joan Mitchell Foundations for Warhol).
Lynn Goldsmith’s 1981 photo of Prince (left) and Andy Warhol’s 1984 artwork (right). Image: United States Supreme Court.
However, the Court has also endeavored to assure us that the disputed use at issue was not Warhol’s creation of the additional fifteen works, but rather the later use of his Foundation. commercial Licence of one of these works at Condé Nast in 2016. Hence Gopnik’s cautious optimism. Maybe then all that appropriation art of the last fifty years doesn’t have a copyright issue, at least as long as it’s not allowed to be used in the tribute issues of glossy magazines. ?
If this all seems a little weird to you, great. It should. This does not correspond to artistic practice. There is no art school that teaches artists to create works for unique, defined, hermetically sealed, attorney-approved uses. Only in the rarest of cases does a creator make something for the limited purpose of donating it to a non-profit museum and pledge never to make any money licensing the copyright. ‘author. Visual artists create works knowing, and often intending, that countless other things will happen to them downstream. They can be sold, hung, resold, given away, loaned, licensed, recast in new forms, critiqued, honored, reissued, parodied, published, bequeathed, featured in a movie, sucked into a machine learning dataset, repackaged in an NFT, etc. Artists make art, not use cases.
Enough reading, the majority opinion actually provides little comfort as to the legality of Warhol’s initial creation of orange prince and other unlicensed works. The Court limited its opinion to the licensing practices of the Warhol Foundation in the 21st century solely because of quirks unique to the litigation. Goldsmith, for example, strategically waived its right to pursue other uses. The Court’s decision not to discuss these cases says nothing about whether they are infringing.
What we got is a new, broadly applicable and transformative fair use test: a test that should concern even artists who have little interest in licensing their creations to magazines. Under the Warhol framework, if you are a practicing artist using copyrighted source works for a purpose “very similar” to the purpose for which the original work was intended, and if you make money from your practice, the onus will be on you if you want to claim transformative fair use under what is known as factor onewhich asks courts to consider the “purpose and character” of the alleged fair use.
What does the Court mean by “highly similar” purposes? This needs to be fixed, but we know how it feels not means: aesthetic, formal, stylistic or semiotic changes are not different objectives. Take a famous example of the kind of appropriation: that of Sherrie Levine After Walker Evans: 4, 1981, a photograph of a reproduction of Walker Evans’ Depression-era portrait of Allie Mae Burroughs. The Metropolitan Museum of Art’s label describes Levine’s aim as a “feminist diversion of patriarchal authority, a critique of the commodification of art, and an elegy on the death of modernism.” But as I read Warhol, this is not the kind of objective that the Supreme Court is asking for. Instead, a functional analysis of the type used by the Court might conclude that Walker Evans (the original photographer) and Levine (the appropriator) share a common goal of portraying literally the same person, Allie Mae Burroughs, in a documentary fashion.
There is no art school that teaches artists to create works for unique, defined, hermetically sealed, attorney-approved uses.
Yes, the Court concedes that changes in meaning or message may still have some relevance in fair dealing investigations. But assuming the artist sells their works for cash, the damage is already done due to the burden they have been burdened with by overlapping goals. The only concrete example we get for what might overcome this burden is when the artist can show that they had to “target” or use the source work, for example to directly parody it. But, to stick to Levine’s example, despite all the richness of her text, is it fair to say that she was parody Walker Evans? No.
What about artists working with generative AI? Could the fair use case be stronger because the purposes diverge more substantially? This argument could hold water when used by AI developers, like Stability AI, seeking to champion their license-free use of large numbers of copyrighted works as training inputs for their AI models. The purpose for which the AI platforms place these images – what a lawyer might call teaching the program how to make art – is apparently different from the purpose of, say, a photographer who sold an original photograph to Getty Images for attribution to others. This result may at first glance seem surprising in a case purported to curtail transformative fair use, as it could make it harder for original content creators to attack AI platforms by invoking fair use as a defense. But it underscores Justice Kagan’s point in footnote five of her dissent that this reductive functional analysis could, counterintuitively, make fair dealing claims easier to defend when the immediate objectives of the content owner (the photographer) and appropriator (the AI developer) diverge even superficially. path.
This same argument, however, has less impact in fair dealing cases where the defendant is an artist. using AI to create new works (as opposed to the platform itself). Now, the copyright holder could convince the court that the relevant goal is no longer to train the AI, but rather to use the AI as a tool to produce a new image for commercial purposes. With this framing, the respective goals of the original photographer and the later artist can seem very similar (to create aesthetically pleasing, licensed visual art).
All of this underscores a particularly fascinating aspect of the competing opinions displayed in Warhol. I am convinced that every judge, if asked, would declare that they ruled as they did in order to support art and artists. They just have very different views on how to shape fair dealing to achieve that common goal.
The end result is that artists who make money from their practices and want to use copyrighted source works must now be prepared to demonstrate, in a way that a federal judge will immediately seize, that their choices to use these particular sources were “reasonably necessary”. to achieve clear goals. If you’re worried that this might force artists to make aesthetic choices they didn’t want to make, that’s understandable.
A risk-averse artist can always, of course, pay to take a license from the creator of the source material, or use their own source photos. It is no coincidence that this is the practice Warhol relied heavily on later in his career, after growing weary of lawsuits. Indeed, if I read the opinion correctly, this is the outcome the current Court most hopes for – a reinvigorated licensing market for source works used in follow-on artistic practices.
The enduring irony, then, is perhaps that Warhol – an artist synonymous with appropriation – may now be the name lawyers cite most to attack this very practice.
The information provided in this article is my opinion only and does not constitute and is not intended to constitute legal advice.
Peter J. Karol is Associate Dean and Professor of Law at New England Law | Boston, where he focuses his research on art and intellectual property law.
