If you’re an artist, creative person, or content creator (and who isn’t these days?), or you work on the creative side of brand marketing, you probably have some pre-existing assumptions about whether and when you can use other people’s work in your creations. Whether it’s Andy Warhol’s soup cans or a sample from a Queen song in Ice Ice Baby, referring to other people’s work in making new creations has always been part of the creative landscape. The question, however, is what is permitted as a “fair use” today, as opposed to what was recognized as fair use in the past.

Some recent shots across the legal bow have brought long-held assumptions into question, and it may be time to rethink our approach to freedom of expression and artistic relevance in the digital age.

MetaBirkins and Artistic Relevance

In one notable case, a jury recently found that an artist who had created digital artworks based on the iconic Hermès Birkin bag was liable for trademark infringement, trademark dilution, and cybersquatting. Mason Rothschild (born Sonny Estival) claimed that his so-called “MetaBirkin” non-fungible tokens (NFTs), which were inspired by the ultra-luxe handbags made famous in Sex and the City, were works of art. The digital works, available in a wide variety of colors and virtual materials, were playful abstractions of the bags, supposedly meant to critique the world of high fashion. Rothschild further claimed that “the First Amendment gives me the right to make and sell art that depicts Birkin bags”. However, both the jury in the case and federal Judge Jed Rakoff of New York’s Southern District disagreed. They awarded Hermès more than $130,000 in damages, and permanently blocked Rothschild and his associates from minting or selling MetaBirkin NFTs.

It appears that a number of factors came into play in the decision, including whether or not Rothschild’s works represented an artistic expression, whether he had intended to mislead consumers into believing his NFTs were backed by Hermès, and whether or not he had acted in good faith when creating them. According to Rakoff, “Hermès proved that Rothschild intentionally misled consumers into believing that Hermès was backing its products.” But just as importantly, Rakoff questioned the artistic intent of the NFTs, opining that “the jury found that Rothschild was simply a swindler.” This was a significant finding, because Rakoff mentioned that had the jury found any indication that there was true artistic intent or expression in Rothschild’s work, he might have been entitled to First Amendment protections. In the end, the judge believed that Rothschild was just as tempted by the fashionably high prices fetched by Birkin bags as he was critical of them.

While NFTs aren’t as hot a commodity today as they were in 2021 when the suit was brought, the decision may well encourage other brands to extend their claims of ownership in the real world by coming after artists who work in the digital world. It seems that artists may be asked to work harder to prove that there is actual artistic merit to their digital interpretations of others’ works rather than just an opportunistic hunger to cash in on digital opportunities. Courts typically avoid being in the position to determine what is art and what is just grifting, but sometimes that role becomes unavoidable. And circumstantial evidence of bad faith intent may be enough to nullify an artist’s claim that a work is subject to the First Amendment right to freedom of expression.

Why Now? The Rogers Test Gets Tested

One of the legal precedents that artist Mason Rothschild was counting on when he went to court to defend himself against Hermès is a case called Rogers v. Grimaldi. It is one of the most notable modern rulings on the balance between trademark rights and free expression. In that case, which was decided in 1989, noted actress and dancer Ginger Rogers sued film producer Alberto Grimaldi and production company MGM over the reference to her name in the title of the Federico Fellini film, Ginger and Fred. Rogers claimed that the film, about two Italian cabaret performers whose act imitated Rogers’ famous pairing with Fred Astaire, violated her trademark rights, and could mislead the public into thinking Rogers herself was involved.

According to the court at the time, the case presented “a conflict between Rogers' right to protect her celebrated name and the right of others to express themselves freely in their own artistic work.” However, the court ruled in favor of Grimaldi, saying that despite its title, suppressing a piece of art like the film would "unduly restrict expression." They also said that the trademark protections of the Lanham Act don’t prevent the use of a celebrity's name in the title of an artistic work “where the title does not explicitly denote authorship, sponsorship, or endorsement by the celebrity or explicitly mislead as to content.”

More importantly, the case led to the common use in legal circles of what is now referred to as “The Rogerstest”. The Rogers test raised the bar for potential trademark infringement plaintiffs, by asking two things about what the artist did:

If both of those two conditions were met, then the court would have to determine if there was a likelihood of confusion and rule accordingly.

While there have been a wide range of interpretations of some aspects of the Rogers ruling over the past few decades, it has generally done a pretty good job of protecting artists ever since. In fact, the basic pillars of the case have often been expanded, such as by including protection for creative commercial works in addition to purely artistic works, and by sometimes including protections for artistic content, rather than just titles as in the original ruling. The test has been applied numerous times in determining the validity of a trademark infringement case against an artist, and often resulted in rulings in favor of the artist. That is, until recently.

The Order of the Orders

So why did things change in the case of the MetaBirkins? It’s not entirely clear, but one aspect, at least according to Rothschild, seems to be the order in which the jury was asked to consider various factors in the case. If you think back to grade school mathematics, you may remember that the order of operations can greatly affect the outcome of an equation. And law is no different. When giving instructions to the jury, the court asked them to first consider whether Rothschild was liable for trademark infringement, trademark dilution, or cybersquatting before even considering whether Rothschild should be legally protected by his right of artistic expression. This marked a significant change from most cases of this nature, which usually require that the jury apply the Rogers test first to determine if the work meets the low threshold for having artistic merit. Typically, the second Rogers question would then be applied to determine if the use was explicitly misleading. Only after satisfying both of those steps would the analysis then turn to whether, even if there was no intent to mislead consumers, such confusion was likely to result anyway.

But the court in Hermès unapologetically rejected the typical procedure. Instead, it instructed the jury to decide on the issues of infringement, dilution, and cybersquatting first before turning to artistic merit as a possible defense. Though the judge stated otherwise, Rothschild asserted that this is may have had some impact on the jury’s decision. Rothschild suggested that once the jury had decided that confusion was likely, they may have been less inclined to think of his works as art and expression, and less likely to excuse him from liability, even if the judge paid lip service to the artistic merit of the works.

In the Hermès case, the court seemed to disregard the struggle between the commercial use of the trademark and its artistic relevance, and expressed confidence in how it had applied the Rogers test in weighing freedom of expression against trademark protection. The court went even further to state that whether “the Rogers test even properly applies to a case like this one has now been cast in doubt.” It characterized Rothschild’s references to the Hermès Birkin trademarks as “explicit and central to Rothschild’s venture.”

But is that a bad thing? How can an artist critique or refer to something clearly without mentioning it explicitly? And how explicit is it if the artist pins additional language like “Meta” onto the name to show that he is referring to it rather than copying it? Does an explicit reference always mean that its use is “explicitly misleading”? This wouldn’t bode well for Warhol’s soup cans. If the “explicitly misleading” aspect of the Rogers test isn’t only limited to how the trademark is used, but also applies to what a jury guesses about the artist’s intent, then this could have a chilling effect on the protections of the First Amendment.

This lack of clarity means that courts now have greater power in determining when expressive freedom should and shouldn’t outweigh likely trademark confusion. That in turn makes outcomes less predictable and could stifle artistic expression, since an artist’s intention in referring to a third-party trademark could potentially be seen as “explicitly misleading” in any context. At the very least, this broad interpretation of the second question of the Rogers test strongly suggests that artists who refer to a known brand in their work need to be very careful in any statements they make about either the reference product or their own.

The Rogers test, which once offered broad First Amendment protections to artists, now seems at risk of becoming solely about guessing the defendant’s intent. This evolution makes it much more likely that courts will circumstantially assume an intention to take advantage of a brand’s equity rather than protect an artist’s right to commentary and free expression. It also puts courts in a dangerous position where they can create a very narrow definition of what is art and what is not, even if they are aware that they are not supposed to make aesthetic judgments.

While it’s important to protect brands from bad faith actors who seek to intentionally confuse consumers, it’s also important to preserve a definition of what is “explicitly misleading” that allows artists to freely comment on society. Courts could better achieve that balance between freedom of expression and trademark infringement by requiring that plaintiffs show actual confusion with the defendant’s product by a significant proportion of consumers (which Hermes purportedly did), rather than just the potential for confusion. In other words, if the Rogers test is to remain useful as a legal standard, it should not be something that courts only tip their caps to before they determine that artistic expression is nonetheless infringing. The Rogers test can only protect First Amendment expression if we require not only that the creative use is “explicitly misleading”, but also that real damage has been done because a substantial proportion of consumers have, in fact, been misled. To do otherwise makes the restrictions on fair use seems quite unfair.