What do a bunch of puppies, a pretty woman, a dancing baby, Demi Moore, some Rastafarians, and 20 million books all have in common? They all refer to prominent, copyright-related cases* from which a content creator could—if he has nothing better to do—learn something about fair use doctrine. But even if an independent artist were to study Rogers v Koons (1992) right through the most recent ruling by the 2nd Circuit Court of Appeals in Google v Authors Guild, I suggest that what he is most likely to discover is that fair use is a highly subjective component of U.S. copyright law. In fact, the subjectivity is part of what makes the principle, and the cases in which it is argued, so interesting—at least to me.
As many readers know, fair use was codified into law in 1976. Functioning as a limitation to copyright, its initial intent was to allow the use of a protected work, regardless of the author’s permission, for the purpose of commenting in some manner on the work itself. Although the doctrine has evolved to include uses that transform works for purposes other than commentary (e.g. Google Books), the underlying need for the fair use limitation was to maintain balance between free speech and copyright. If an author were able to enforce his copyright in order to stifle a use necessary to comment on his expression or that transforms his expression in some useful way, then free speech may be chilled or copyright may fail in its purpose to promote expression or progress. But drawing these contours is, of course, only possible on a case-by-case basis. When a defendant argues fair use in a copyright infringement suit, a panel of judges will consider these four factors:
- The purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit purposes.
- The nature of the copyrighted work.
- The amount and substantiality of the portion used in relation to the copyrighted work as a whole and relative to the purpose of the new work or use.
- The effect of the use upon the potential market for the value of the copyrighted work.
Even a casual glance at these criteria should tell independent creators what their attorneys would tell them, if they had attorneys—that there are no bright lines in fair use defenses. Each case is distinct and highly susceptible to human interpretation. For example, I bet most laymen—me included—would have a tough time squaring the 1992 decision rejecting artist Jeff Koons’s fair use defense for his use of Art Rogers’s “Puppies” photograph with the 2013 decision affirming Richard Prince’s fair use defense for his use of Patrick Cariou’s Rastafarian photos. The reason I cite these two cases is that when considering the fourth factor in the latter, the panel determined that because Prince’s market is the fine art world and Cariou’s market is not, the potential market harm to the original photographs would be unlikely. Aside from the fact that this contemporary ruling places judges in the undesirable role of cultural critics, the point I’m making is that the same distinction could have been made between Jeff Koons and Art Rogers, but it was not.
Forgive that detour, but I’m trying to underscore that, although there are some fairly simple circumstances in which a rights holder might make a reasonable assessment about a fair use of her work, case law demonstrates that there is nothing universally objective about the doctrine—perhaps especially when the use is artist to artist. In fact, my colleague Terry Hart tells me that between 1978 and 2005, 33.8% of District Court fair use decisions were overturned on appeal. And there’s a reason why the landmark case involving 2 Live Crew’s parodic version of the song “Pretty Woman” went all the way to the Supreme Court: because fair use is complicated. So, the suggestion that a non-attorney, creator can be expected to make an “objective” assessment of a possible fair use defense prior to sending a DMCA notice to take down an otherwise infringing use of her work online is an exercise in metaphysics, if not outright sorcery.
My apologies at this point for the amount of amateur legal analysis in this piece—something I normally avoid—but there’s no other way to tell this story.
The reason I’m stressing the subjective nature of fair use is that this week an amicus brief was filed on behalf of Google, Twitter, WordPress, and Tumblr asking the Court of Appeals for the 9th Circuit to overturn—not its decision—but its rationale in the 2004 case Rossi v MPAA. And it appears, based on the number of references to fair use in the brief, that they are seeking this revision because it might strengthen the Internet industry’s position regarding the ruling just last month in Lenz v UMG, a.k.a. the “Dancing Baby” case. (This involves the temporary takedown from YouTube of a home video depicting Ms. Lenz’s baby dancing to Prince’s “Let’s Go Crazy,” which occurred in 2007 and has been litigated by the EFF ever since.)
The recent Lenz ruling by this same court affirmed that a rights holder, prior to sending a DMCA takedown notice, must consider whether or not a use would be judged fair but that the consideration may be “subjective,” thereby rejecting the EFF’s argument in that case for an “objective” standard. So, it’s interesting that the Internet companies named in this brief are seeking a revision to the rationale in Rossi, in which the court also applied a “subjective” standard, but in a lawsuit that had nothing whatsoever to do with a fair use defense. So, why seek the new ruling in Rossi instead of Lenz? Because in a weird way, it might work. But not necessarily.
In Rossi v MPAA, Michael Rossi claimed that the Motion Picture Association wrongfully used the DMCA to shut down his site internetmovies.com. The takedown was issued because the site had advertised that full-length movies could be downloaded from the portal, though this turned out not to be true. Rossi was lying to users (presumably in order to generate traffic), and the MPAA did not confirm whether or not his platform actually made movies available before they sent the takedown notice. As a result, Rossi sued MPAA for a variety of injuries stemming from its alleged abuse of DMCA, but the court held that MPAA had acted in good faith in accordance with the DMCA and so rejected all of Rossi’s claims.
This new amicus brief filed on behalf of Google et al, states that the outcome of Rossi is correct but that the court should “take this opportunity to clarify the law and hold that that the good faith requirement in Section 512(c)(3)(A)(v) encompasses an objective standard with respect to whether use of a copyrighted work is ‘authorized by law.’” In theory, if these Internet companies can convince the court to affirm an “objective” standard in Rossi, the revision could perhaps be applied to the decision in Lenz and then more broadly assert that any rights holder must “objectively” consider fair use before issuing a takedown notice, or risk possible litigation for wrongful takedown.
But even if the court were to revise its ruling to affirm an “objective” standard in Rossi, any reasonable person should recognize that the additional burden on the MPAA in that case would be a consideration that is, in principle, objectively possible (i.e. they might have been able to confirm whether or not Rossi’s site was actually making movies available illegally). But, in a given situation in which a rights holder contemplates issuing a takedown notice, there are almost no objective criteria (i.e. evidence) by which he can consider whether or not the recipient of the notice might present an effective fair use defense. (As addressed in this post about an artist friend of mine, fair use confuses people with the best intentions.) The dynamic and interdependent nature of the four factors alone should demonstrate to any reasonable individual that fair use is a highly subjective, case-by-case doctrine.
More specifically, the “subjective” ruling in Lenz refers to the automated notice and takedown procedure, typically used by major rights holders like UMG, coupled with an algorithmic (i.e. imperfect) process to assess likely fair uses. The court held that this “subjective” automation meets the standard of a good faith effort, which seems only reasonable in light of the tens of millions of rightful DMCA takedown notices that are sent out by major rights holders every month. And that brings us to the heart of the matter …
Can we address takedown abuse without improperly burdening rights holders?
Ostensibly, the aim of this amicus brief is to address the problem of abusive takedown; and the authors state their broad concern about the “subjective” standard thus:
“… the more misinformed or unreasonable the copyright owner, the broader the immunity he would have from liability under Section 512(f). This reading of 512(f) would effectively encourage copyright owners to remain ignorant about the limitations on their exclusive rights under the Copyright Act, see 17 U.S.C. §§ 107–123, because the less they know, the more leeway they would have to send takedown notices.”
Perhaps there is some merit to this anxiety, but I am doubtful that the solution will be found in seeking the “objective” standard revision in Rossi. After all, the takedown of Ms. Lenz’s video did not occur due to ignorance of the law; the video was restored to YouTube according to DMCA procedures; and it was the EFF that decided to spend the last eight years suing UMG because the case appears to have provided a pathway to reshape the law behind the PR veil of “big mean corporate rights holder picking on an innocent baby.” As I’ve said in the past, there are better examples of DMCA takedown abuse, but not so many with headliner names like Prince.
No question that DMCA takedown abuse does happen, though it is dramatically outweighed by the number of legitimate takedown notices that are sent and resent in a nearly futile attempt to stop countless incidents of actual infringement. The amicus brief cites several exemplary takedown abuses—many of which would likely merit a fair use defense—issued by individuals or entities who either don’t understand copyright or who know exactly what they’re doing and are abusing copyright in an attempt to censor criticism or commentary. But these examples are not an indictment of the purpose of copyright; and rights holders with legitimate claims should not be made to bear the burden of mitigating abuse by a minority of bad actors—least of all with the purpose of making DMCA procedures easier or cheaper for OSPs. In fact, the weakness of the stated motivation for this amicus brief is written in the document itself and stated thus:
“Google receives hundreds of notices that suffer from similar defects, often repeatedly from the same vexatious submitters, and devotes substantial human and machine resources in an attempt to identify these abusive notices among the tens of millions of DMCA notices that Google processes each month”
Out of tens of millions of notices, Google identifies hundreds of potentially abusive ones from a consistent group of “vexatious submitters.” Does that not sound on the face of it like an exception to a rule, one that begs for a targeted solution rather than a broad revision of legal standards that may place undue burden on many rights holders? The Internet industry regularly criticizes proposals for remedies and legal frameworks for being “overly broad.” Yet, the pursuit of this revised standard, seems to be asking the court to use a sledgehammer to swat at flies. Hundreds of abusive notices out of tens of millions is somewhere in the universe of .003%, and the Internet industry wants us to believe that this is the real problem with DMCA.
It seems to me that the public interest would be best served by Congressional revision of the DMCA in order to mitigate both takedown abuse and safe harbor abuse (though you may hear a collective gasp from Silicon Valley at the suggestion of the latter). In fact the amount of energy and resources the Internet industry has devoted to rewriting Lenz into a story of “abuse” is both revealing and appalling. After all, it is unreasonable to argue that a rights holding entity must be held to a stricter standard of consideration of the legality of a particular use while inadequacies in the antiquated safe harbor provisions necessitate the automated issuing of tens of millions of takedown notices per month. In other words, if Google et al want rights holders to more carefully consider each infringement, then Google et al should be required to more aggressively reduce the volume of infringement on their platforms down to a manageable scale. Meanwhile, I believe neither the courts nor Congress should amend standards that can in any way increase the challenges already faced by individual and small-entity creators to enforce their copyrights in the digital marketplace. In fact, making that enforcement easier just might benefit everyone.
* In order as referenced: Rogers v Koons; Campbell v Acuff-Rose; Lenz v UMG; Leibovitz v Paramount Pictures; Prince v Cariou; Google v Authors Guild.