This lawsuit stems from the Google Book Project, an ambitious program launched by Google in 2004 to digitize the library collections of the University of Michigan, Harvard, Stanford, the University of Oxford and the New York Public Library and make the collections available for searching online. In 2008, a group of universities established HathiTrust as a repository to combine, archive and share their digital libraries, and make the collection available to the public. At the time of the commencement of suit, the repository reportedly contained 10 million volumes. The Authors’ Guild, an authors’ trade association, and other authors’ groups sued HathiTrust for copyright infringement and swiftly moved for a judgment on the pleadings that HathiTrust could not rely on the fair use defense to the claim of copyright infringement.
First the Court addressed the issue whether the fair use defense is available to library institutions, or whether such institutions are limited to the separate defense to infringement found in Section 108 of the Copyright Act and known as the “library exception.” The court concluded that “fair use does not undermine Section 108, but rather supplements it.” The court then addressed the four fair use factors: purpose and character of the use; nature of the copyrighted work; amount and substantiality of the portion taken in relation to the copyrighted work as a whole; and the effect on the market for or value of the copyrighted works. The court found:
1. Purpose and character of the use. The court found that the use that HathiTrust is making of the works is transformative because it serves an “entirely different purpose than the original works.” The purposed is “superior search capabilities rather than actual access to copyrighted material.”
2. Nature of the copyrighted work; amount and substantiality of the portion used. The second and third factors were not considered important, though the court rejected the plaintiffs’ contention that the uses could not be fair because they made use of the entire works at issue.
3. Market effect. The court found that this factor weighed in favor of fair use for a number of reasons. The plaintiffs could not identify any specific quantifiable harm to the market to exploit their works, or any documents relating to such harm. The court noted that HathiTrust has implemented security measures in place to prevent wholesale infringing copying. Finally, the transformative nature of the use undercut any actionable market harm, as a copyright holder cannot preempt a transformative market.
Balancing all the factors, the court found the use fair. The case has been appealed. A more detailed discussion of the Google Book project can be found in my paper, “Fair of Foul? Mass Digitization and the Fair Use Doctrine,” submitted in connection with the AIPLA’s 2012 Spring Meeting and available at http://www.shadesofgraylaw.com/wp-content/uploads/2013/04/AIPLA-Paper-2012.pdfPrint This Post
Last week I had the privilege of addressing the Dallas Bar IP Section on the subject of the current controversy over copyright and prior art submissions in patent prosecution. Four lawsuits have been filed by publisher John Wiley & Sons and the American Institute of Physics against law firms alleging copyright infringement for reproducing and distributing various scientific articles in the course of preparing and submitting patent applications. The lawsuits allege that the law firms violated copyright in the articles at issue by (1) making and distributing copies of the articles to the USPTO in connection with patent applications; (2) making additional copies of articles cited in patent applications; and (3) making copies of articles that they neither cited nor submitted to the USPTO for internal purposes. One of the cases apparently settled over the summer. In two of the remaining cases, the plaintiffs have amended their pleadings to drop the allegations concerning submissions to the USPTO, leaving only the allegations regarding internal law firm copying.
The copyright and patent communities have taken note of these lawsuits, and there is an ongoing discussion regarding whether the practices at issue constitute fair use of the articles in question. In January of 2012, the General Counsel of the USPTO issued a memorandum asserting that the accused practices constitute fair use, and the USPTO has intervened in two of the cases as a defendant and counterclaimant, seeking a declaration of noninfringement. In this post, I will give an overview of the fair use doctrine and apply it to the copying at issue in the prior art cases. Read morePrint This Post
I drafted this piece in preparation for my ABA-sponsored speaking engagement in Washington, D.C. on June 10, 2011. If you’re going to be in the DC area on that day, please consider joining us. In addition to the Vernor case analyzed below, my panel will also address the Ninth Circuit’s opinion in UMG v. Augusto and the future of the first sale doctrine in the wake of these two significant cases. There will also be panels on topics relating to trademarks, patents and trade secrets. To register for the event, click here.
* * * * * * *
The first sale doctrine entitles “the owner of a particular copy . . . lawfully made under this title” to “sell . . . that copy.” 17 U.S.C. § 109(a). The Supreme Court first recognized this doctrine in 1908 in Bobbs-Merrill Co. v. Straus, 210 U.S. 339 (1908). In that case, a book publisher sold copies of a book to the wholesale trade bearing a notice stating that the book could not be sold for a retail price of less than $1. There was no license agreement between the book publisher and wholesale purchasers. The defendants bought copies of the book and resold them at retail for less than $1 per copy. The Supreme Court held that the Copyright Act did “not create the right to impose, by notice . . . a limitation at which the book shall be sold at retail by future purchasers, with whom there is no privity of contract.” Bobbs-Merrill, 210 U.S. at 350. “[O]ne who has sold a copyrighted article, without restriction, has parted with all right to control the sale of it. . . . [T]he books sold by the appellant were sold at wholesale, and purchased by those who made no agreement as to the control of future sales of the book, and took upon themselves no obligation to enforce the notice printed in the book, undertaking to restrict retail sales to a price of one dollar per copy.” Because the Copyright Act did not grant the publisher the right to control the pricing of downstream sales, and there was no privity of contract between the publisher and the defendants, the defendants were free to resell the books as they saw fit.
Over the last 100 years, the first sale doctrine has furthered the dual purposes of the U.S. Constitution’s Copyright Clause by acting as a limitation on the copyright owner’s monopoly and providing a corresponding benefit to the public’s interest in access to creative works. In June 2010, the Ninth Circuit dealt a significant blow to the applicability of the first sale doctrine to software when it ruled that a downstream purchaser had obtained a mere license to the software and could not resell his copy of it. Vernor v. Autodesk, 621 F.3d 1102 (9th Cir. 2010) (“Vernor II”). Timothy Vernor, an eBay reseller, purchased several copies of Autodesk’s AutoCAD software from CTA, one of Autodesk’s direct customers, and sought to resell them on eBay. Vernor succeeded in selling some copies, but after Autodesk served eBay with repeated DMCA takedown notices accusing Vernor of copyright infringement, eBay eventually terminated his account, effectively prohibiting him from selling his remaining copies.
District Court proceedings and opinion
Vernor brought an action for declaratory relief, seeking a declaration that his sales of AutoCAD were noninfringing because he was an “owner” of the “particular copy” of each software package pursuant to §109. Vernor also claimed immunity under §117 of the Copyright Act, which authorizes the “owner of a copy of a computer program” to make a copy of the program if doing so is an “essential step in the utilization of the program.” This provision exempts from liability the copying of a software program into a computer’s memory as part of the program’s normal operation.
Autodesk argued that though it transferred possession of the software to CTA, its license agreement (“License”), which imposed a variety of limits on CTA’s ability to use and dispose of the software, did not transfer ownership of the software to CTA. The License reserved title and copyright in the software to Autodesk. Moreover, the License allowed CTA to install the software on only two computers at a time, and prohibited CTA from using the software simultaneously on those computers. The License forbade CTA from modifying or reverse engineering the software, and from using or transferring the software outside the Western Hemisphere. It likewise barred any transfer of the software without Autodesk’s written permission. And it provided that if CTA obtained the software as an upgrade from an earlier version of the software, CTA must destroy its copies of the earlier version. When CTA purchased the software packages at issue, however, it did not agree to destroy them if it later upgraded to a subsequent version.
The District Court ruled in Vernor’s favor, finding that Autodesk transferred ownership of the AutoCAD packages to CTA, which in turn transferred its ownership interest to Vernor. In so doing, the District Court made an important distinction: that although there was no dispute that Autodesk licensed the software itself to CTA, “the use of software copies can be licensed while the copies themselves are sold.” Vernor v. Autodesk, 2009 U.S. Dist. LEXIS 90906 (W.D. Wa. Sept. 30, 2009)(“Vernor I”).
In order to determine who was the “owner” of the AutoCAD copies, the District Court attempted to reconcile four apparently conflicting Ninth Circuit opinions. United States v. Wise, 550 F.2d 1180 (9th Cir. 1977), involved movie prints distributed by studios to theaters for display, or to various “V.I.P. licensees” like the actress Vanessa Redgrave for personal use. The Wise panel looked to the terms of the applicable transfer agreements to assess whether the studios had transferred ownership of the prints or merely licensed them. The transfer agreements contained a variety of contradictory terms; some suggested a transfer of ownership, while others suggested a mere license. Thus, the Wise panel considered whether an agreement contained a clause reserving title in the transferred work to the copyright owner; whether the transferee made a single, up-front payment; whether the agreement required the transferee to return or destroy the transferred copy after a certain period of time; and whether the agreement imposed restrictions on the use or subsequent disposition of the copy. The Vernor I court found that none of these terms was dispositive of the issue, with one exception: the Wise panel found a transfer of ownership in each agreement which allowed the transferee to retain possession of the transferred copy indefinitely, and gave the copyright owner no right to reclaim the copy. By contrast, the Wise panel construed as a license each agreement which gave the copyright holder the right to regain possession of the transferred copy.
The Vernor I court concluded that, under Wise, Autodesk had transferred ownership in AutoCAD to CTA. Though the License purported to reserve title in the software to Autodesk, and imposed restrictions on the use and disposition of the transferred copy, it did not allow Autodesk to regain possession of the copy. “In [the District Court’s] view, retaining title in a copy is meaningless unless the copyright holder has some means to regain possession of the copy.”
The Vernor I court then turned its attention to a trio of Ninth Circuit opinions issued long after Wise, in the digital era: MAI Sys. Corp. v. Peak Computer, Inc., 991 F.2d 511 (9th Cir. 1993); Triad Sys. Corp. v. Southeastern Express Co., 64 F.3d 1330 (9th Cir. 1995); and Wall Data Inc. v. Los Angeles County Sheriff’s Dep’t, 447 F.3d 769 (9th Cir. 2006). Each of these cases addressed whether a licensee was an owner entitled to invoke the “essential step” defense in §117. In MAI, the Ninth Circuit addressed whether a computer repair service that serviced computer systems containing licensed software made unauthorized copies of that software when they operated the computers that they were servicing. In a footnote, the court noted, “Since MAI licensed its software, [its] customers do not qualify as ‘owners’ of the software and are not eligible for protection under §117.” Consequently, those customers could not authorize the repair service to make repairs to the computers, since doing so necessarily involved the making of a copy of the software in the computer’s memory.
Triad also involved software and a computer repair service. The software in Triad was distributed pursuant to three different agreements. The first sold the software outright; these transferees were owners under §117. The second licensed the software but restricted duplication and third-party use, and the third added a requirement that licensees pay a transfer fee to sell computer systems on which the software had been installed. Transferees under these agreements were licensees, not owners, and could not authorize repairs pursuant to §117.
Finally, in Wall Data, the Los Angeles County Sheriff’s Department purchased software pursuant to a license which (1) prohibited the department from installing the software in “multiple computer” or networked arrangements; (2) restricted transfers of the software between computers to once every thirty days; and (3) limited the department to installing the software on 3600 machines. The license did not, however, prohibit resale of the software. The department proceeded to install the software on more than 6000 machines, but configured it so that no more than 3600 users could access it at any given time. The Wall Data panel held that “if the copyright owner makes it clear that she or he is granting only a license to the copy of software and imposes significant restrictions on the purchaser’s ability to redistribute or transfer that copy, the purchaser is considered a licensee, not an owner, of the software.” As in MAI, the Wall Data licensing agreement “imposed severe restrictions on the Sheriff’s Department’s rights with respect to the software. Such restrictions would not be imposed on a party who owned the software.” Thus, the Sheriff’s Department did not own its copies of the software, and could not rely upon the essential step defense. The panel also denied §117 protection to the Sheriff’s Department because copying the software to the additional computers was not an “essential step” in the operation of the software.
The Vernor I court concluded that if it were to follow the MAI trio, “Autodesk would prevail.” Under MAI and Triad, “the mere labeling of an agreement as a license is sufficient to ensure that the licensee does not have ownership of any copy of the software.” The Wall Data holding, while “more flexible” according to the Vernor I court, nonetheless provided that “if the copyright owner makes it clear that she or he is granting only a license to the copy of software and imposes significant restrictions on the purchaser’s ability to redistribute or transfer that copy, the purchaser is considered a licensee, not an owner, of the software.”
The Vernor I court found Wise and the MAI trio in irreconcilable conflict with each other. Under the reasoning of Wise, Vernor should prevail, because the Autodesk License allowed CTA to retain possession of the copy of the software indefinitely. Under the MAI trio, Autodesk should prevail, because the licenses in question unequivocally provided that they were licenses, and reserved ownership to the copyright holder. When precedent conflicts, the court must follow “the oldest precedent among conflicting opinions from three-judge Ninth Circuit panels.” Consequently, the Vernor I court applied Wise and ruled in favor of Vernor.
Ninth Circuit opinion
The Ninth Circuit reversed, finding that Autodesk’s direct customers were licensees, not owners, and therefore could not transfer ownership rights to Vernor. The Vernor II court found no conflict between Wise and the MAI trio. It reasoned that under Wise, the Ninth Circuit considers all of the provisions of a transfer agreement to determine whether it constitutes a license or a transfer of ownership, including “(1) whether the agreement was labeled a license and (2) whether the copyright owner retained title to the copy, required its return or destruction, forbade its duplication, or required the transferee to maintain possession of the copy for the agreement’s duration.” Vernor II rejected Vernor I’s interpretation of Wise as holding that “a transferee’s right to indefinite possession itself established a first sale.” To the contrary, no one factor is dispositive, and each transfer agreement must be examined in its entirety.
The Vernor II court characterized the MAI and Triad agreements as “restrictive license agreements,” and thus their customers were licensees who could not qualify for the essential step defense. Finally, the Vernor II court interpreted Wall Data as holding “that the essential step defense does not apply where the copyright owner grants the user a license and significantly restricts the user’s ability to transfer the software.” Thus, the Vernor II court read “Wise and the MAI trio to prescribe three considerations that we may use to determine whether a software user is a licensee, rather than an owner of a copy. First, we consider whether the copyright owner specifies that a user is granted a license. Second, we consider whether the copyright owner significantly restricts the user’s ability to transfer the software. Finally, we consider whether the copyright owner imposes notable use restrictions.”
The Vernor II court went on to hold that “a software user is a licensee rather than an owner of a copy where the copyright owner (1) specifies that the user is granted a license; (2) significantly restricts the user’s ability to transfer the software; and (3) imposes notable use restrictions.” It found that Autodesk’s agreement was a license rather than a sale. “Autodesk retained title to the software and imposed significant transfer restrictions: it stated that the license is nontransferable, the software could not be transferred or leased without Autodesk’s written consent, and the software could not be transferred out of the Western Hemisphere.” Moreover, the License “imposed use restrictions against the use of the software outside the Western Hemisphere and against modifying, translating, or reverse-engineering the software.” The License provided for termination “upon the licensee’s unauthorized copying or failure to comply with other license restrictions. Thus, because Autodesk reserved title to [AutoCAD] copies and imposed significant transfer and use restrictions, we conclude that its customers are licensees of their copies . . . rather than owners.” As licensees, Autodesk’s direct customer thus could not transfer ownership to Vernor. In turn, Vernor could not invoke the first sale doctrine or the essential step defense.
Vernor II raises the question whether and to what extent the first sale doctrine will retain its vitality in the digital age. It vests software developers with tremendous power to restrict downstream uses of their products. Purchasers of software rarely – if ever – have the opportunity to negotiate the licenses that come with the products they think they are “purchasing.” Under Vernor II, saying it makes it so – the opinion makes it all too easy for owners of all types of digital content to include restrictive language in their licenses and strip §109 of any applicability to digital technologies. Nor is it too great a stretch to imagine Vernor II being extended to other, traditional, contexts. Imagine, for example, a print book purchased pursuant to a shrink-wrap license that prohibits the purchaser from reselling it on eBay or outside the United States.
Vernor II also appears to conflate the important distinction between a copyrighted work and the tangible medium in which it is reproduced. In Vernor, the copyrighted work was the AutoCAD software program, but that program was reproduced and distributed on a physical object – a disc. “Ownership of a copyright . . . is distinct from ownership of any material object in which the work is embodied.” 17 U.S.C. § 202. Transferring a copy does not necessarily transfer the copyright, and vice versa. As the Vernor I court noted, “the use of software copies can be licensed while the copies themselves are sold.” Yet nowhere did the Vernor II court explicitly address the issue of ownership of the physical discs as distinct from the copyrighted AutoCAD program. Consider a printed book: the purchaser obtains no rights in the words printed on the pages of the book (the intellectual property), but the purchaser plainly owns the paper upon which the words are printed. The same rationale can be used to distinguish software code from the disc on which it is embedded. Of course, as commerce and digital products increasingly move into the cloud, where products are simply downloaded rather than distributed in a physical medium, this distinction may likewise lose relevance.
On the other hand, the software industry has built itself and its various business models on the foundation that software is licensed, not sold. A ruling in favor of Vernor could upset settled expectations in an industry with a significant impact on the economy.
Vernor’s deadline to seek certiorari to the Supreme Court was May 18, 2011. It remains to be seen whether he can defy the very slim odds of winning Supreme Court review. Until then, the continuing applicability of the first sale doctrine in the context of digital technologies will remain in flux.Print This Post
Wondering whether your contemplated use of copyrighted material is a fair use? Perplexed by the thicket of seemingly contradictory opinions on the four statutory fair use factors? Eager to learn more about transformative uses? Join me on July 12, 2010 for an informative presentation on fair use as part of PLI’s program Understanding Copyright Law 2010.Print This Post
Joel Tenenbaum rang in the New Year by seeking a new trial or, alternatively, a reduction of the $675,000 award of statutory damages imposed by a jury last summer for his illegal file-sharing activities. Attempting to capitalize on favorable language in the court’s opinion rejecting his fair use defense, Tenenbaum’s motion trumpeted the court for becoming “the first to recognize a fair use interregnum for copyright infringement following the debut of Napster.” Tenenbaum can’t be blamed for trying to turn lemons into lemonade, but the court recognized no such thing, though it plainly wished to rule otherwise. At most, the court speculated that a file sharer “might” be able to rely on a fair use defense under certain limited circumstances not applicable to Tenenbaum’s conduct, such as by swapping files during the time period “before digital media could be purchased legally, but  later shift[ing] to paid outlets.”
Tenenbaum devoted much of his brief to rehashing the same fair use arguments that the court already (properly) rejected. After praising the court for supposedly establishing the so-called “fair use interregnum,” Tenenbaum faulted the court for cutting it off before his file-sharing was detected in August 2004 due to the fact that “a commercial market for digital music had fully materialized” by then. According to Tenenbaum, early online sellers of digital music did nothing to alleviate the injustice of having to purchase entire CD’s rather than individual songs because they employed encryption technology which restricted purchasers’ ability to transfer songs between different media players. This “boxed music consumers like Tenenbaum into an unfair choice” until the music industry began offering unrestricted copies of songs for sale online in 2007. Tenenbaum also reasserted his “attractive nuisance” argument – that the music industry lured Tenenbaum and other consumers into wrongdoing with their marketing strategies – and bemoaned anew the conscription of parents and universities as “copyright police to regulate internet use” by children and students. None of these arguments is likely to persuade the court that it made a mistake in rejecting Tenenbaum’s fair use defense, which remains, as the court succinctly noted, “completely elastic, utterly standardless, and wholly without support.”
Statutory damages and due process
Tenenbaum buried his most interesting, and potentially most successful, argument at the end of the brief – that the $675,000 statutory damages award is so grossly excessive that it violates his Constitutional right to due process. The Supreme Court has held that statutory damages violate due process where they are “so severe and oppressive as to be wholly disproportioned to the offense and obviously unreasonable.” St. Louis, I.M. & S. Ry. Co. v. Williams, 251 U.S. 63, 67-68 (1919). In Williams, the Court upheld a $75 statutory damages award against a railroad that had overcharged passengers by 66 cents per ticket. Relying on the Williams standard, Tenenbaum contrasted the “bankrupting” size of the award with his conduct, which he characterized as “at most comparable to shoplifting music from a record store.” Assuming a purchase price of 99 cents per song, Tenenbaum calculated that the ratio between the award and the actual damage to the plaintiffs was 22,500 to 1 – far in excess of the 113 to 1 ratio which the Williams Court found acceptable.
The shoplifting analogy is a compelling one. The consequences for theft can be severe, but it is almost inconceivable that Tenenbaum would have been assessed a six-figure penalty for walking out of a record store with a handful of CD’s stuffed under his jacket. The methodology he used to arrive at his 22,500:1 ratio is flawed in that it fails to account for the fact that a single song, made available over a peer-to-peer system, could be copied innumerable times, thus resulting in more than one lost sale to the plaintiffs. Nonetheless, the award is strikingly high given the nature of the offense. The judge has already expressed considerable sympathy for Tenenbaum and distaste for the record industry’s strategy of suing individual file sharers; the argument that a penniless college student should not be bankrupted for a relatively petty offense could well resonate with the court.
Tenenbaum also argued that the damages award violated due process under the standard applied to punitive damages. A punitive damages award which is “grossly excessive” in relation to the state’s interest in punishment and deterrence enters “the zone of arbitrariness that violates the Due Process Clause of the Fourteenth Amendment.” BMW v. Gore, 517 U.S. 559, 568 (1996). Courts examine three “guideposts” to determine whether punitive damages are appropriate: (1) the degree of reprehensibility of the defendant’s conduct; (2) the disparity between the actual or potential harm suffered by the plaintiff and the damages award; and (3) the difference between the award and the civil penalties authorized or imposed in comparable cases. Id. at 575. Tenenbaum contended that his conduct was not reprehensible, involving only economic harm not motivated by intentional malice and conduct that “even now many see as having been unauthorized but not morally wrong”; that the Supreme Court has noted that “few awards exceeding a single-digit ratio between punitive and compensatory damages, to a significant degree, will satisfy due process”; and that statutory damage awards in file-sharing cases in which the defendants lost by default or on summary judgment have been limited to the minimum possible statutory damage amount.
In citing Williams as the standard for assessing the appropriateness of a statutory damages award in the copyright context, and addressing Gore as an alternative argument, Tenenbaum departed from the post-trial strategy employed by Jammie Thomas-Rasset, the other file sharer to seek modification of a massive statutory damages award. Thomas-Rasset was found liable for copyright infringement for sharing 24 songs, and a jury initially awarded the plaintiffs $222,000. The court vacated that award due to concerns over the propriety of the jury instructions concerning liability. After a second trial, Thomas-Rasset was again found liable, and the jury awarded the plaintiffs $1.9 million in damages. Thomas-Rasset moved for a new trial, relying primarily on the standard set forth in Gore, essentially conflating the two tests (“The Due Process jurisprudence that is today embodied in BMW v. Gore has its roots in Williams, a case involving statutory damages”).
Tenenbaum then argued that if the court does not grant him a new trial, it should reduce the statutory damages award to the minimum amount. Remittitur is appropriate where an award is “grossly excessive, inordinate, shocking to the conscience of the court, or so high that it would be a denial of justice to permit it to stand.” Here, Tenenbaum’s strategy mirrored Jammie Thomas-Rasset’s; she also sought remittitur as an alternative to her due process argument. The court recently granted Thomas-Rasset’s request, reducing the $1.9 million award to $54,000.
Tenenbaum argued that Congress set the currently applicable range of statutory damages to combat large-scale commercial piracy of software over the Internet, but did not intend to subject consumers like Tenenbaum to the upper limit of available damages. According to Tenenbaum, Congress sought to remedy the tremendous costs of software piracy to software companies, their employees and the economy – concerns that sound strikingly familiar in the music file-sharing context – but that it envisioned imposing those damages on those who made the software available for download on a widespread basis and not the individuals who actually downloaded them. Of course, Tenenbaum himself placed songs in the shared folder of his hard drive, making them available for download to anyone within his peer network, so it is unclear how, in practical effect, his conduct differed from that of the software pirates, except perhaps in degree.
Moreover, Tenenbaum offered no justification for why the minimum statutory damages amount is the appropriate award, as opposed to some other amount. This omission underscores the difficulty associated with his request; namely, how the court can defensibly set a damages award within such a large range of potential statutory damages. Perhaps the court will look for guidance to the recent decision in the Thomas-Rasset case; there, the judge settled on a trebling of the minimum award per sound recording as the appropriate amount, based on treble damage provisions in other federal statutes.
The Department of Justice, intervening to defend the constitutionality of the Copyright Act’s statutory damages provisions, and the plaintiffs have submitted their briefs opposing Tenenbaum’s motion. A hearing is scheduled to occur on February 23, 2010.Print This Post
On December 7, 2009, the District of Massachusetts issued a remarkable written opinion elaborating upon its earlier ruling that individual file sharing did not constitute fair use in Sony BMG Music Entm’t v. Tenenbaum, 2009 U.S. Dist. LEXIS 112845. The case stemmed from Boston University graduate student Joel Tenenbaum’s file-sharing activities, which spanned several years and multiple file-sharing services.
The opinion stands out in a number of respects, but most starkly for (1) its eagerness to find any basis to rule in Tenenbaum’s favor and (2) its scathing assessment of defense counsel’s performance. The court was “deeply concerned by the rash of file-sharing lawsuits, the imbalance of resources between the parties, and the upheaval of norms of behavior brought on by the internet,” and did “everything in its power to permit Tenenbaum to make his best case for fair use.” Courts don’t usually go to such lengths to advance one party’s interests, at least not where the party is represented by counsel. Here, Tenenbaum was represented both by a private law firm and by a Harvard Law School professor – a team presumably capable of advancing his interests without an assist from the judge. But this opinion conjures up the image of a judge itching to vault over the bench to argue Tenenbaum’s case for him:
“[T]he court was prepared to consider a more expansive fair use argument than other courts have credited . . . For example, file sharing for the purposes of sampling music prior to purchase or space-shifting to store purchased music more efficiently might offer a compelling case for fair use. Likewise, a defendant who used the new file-sharing networks in the technological interregnum before digital media could be purchased legally, but who later shifted to paid outlets, might also be able to rely on the defense.”
Tenenbaum made none of these arguments, however, and the court deplored – in unusually harsh and explicit terms – his counsel’s performance in the case. Among a litany of other transgressions, the court chastised counsel for litigating the fair use defense “as an afterthought, and literally on the eve of trial,” and characterized the defense as “truly chaotic” and based on “perfunctory” papers. Indeed, Tenenbaum’s papers opposing summary judgment on fair use were structured skeletally, resembling an outline more than a substantive brief; cited only sparsely to the record and to caselaw; addressed the four traditional fair use factors in cursory fashion while emphasizing arguments drawn from unrelated legal doctrines; and invoked generalized incantations of “fairness” more reminiscent of the playground than the courtroom.
Possibly the most interesting insight into Tenenbaum’s defense, however, comes from his own attorney’s legal blog. Following Tenenbaum’s loss at trial, and public criticism of his fair use defense, his attorney, Harvard Law School professor Charles Nesson, solicited feedback in the blogosphere on what alternative defenses commentators felt might have prevailed. In response to one commentator’s list of potentially successful arguments, Mr. Nesson wrote, “these defenses do not join the fundamental issues. this [sic] trial was not an exercise in getting joel off the hook.” The notion that counsel’s job could consist primarily of anything other than exonerating his client should boggle any practicing litigator’s mind. Presumably, Tenenbaum – now saddled with a $675,000 verdict – might wish that his counsel had been more concerned with getting him “off the hook” than with transporting the fair use doctrine to a galaxy far, far away.
Fair Use Analysis
Despite its apparent desire to find in Tenenbaum’s favor, the court correctly noted that the fair use “analysis is not some open-ended referendum on ‘fairness,’ as [Tenenbaum] would have it, but an effort to measure the purpose and effects of a particular use against the incentives for literary and artistic creation that drive copyright protections.” Consistent with mainstream fair use jurisprudence, the court examined each of the four statutory factors and concluded that each one weighed against a finding of fair use. But the court’s overt predisposition in Tenenbaum’s favor unmistakably influenced its reasoning.
For instance, the court refused to “label” Tenenbaum’s conduct as commercial because “there is a meaningful difference between personal file sharing and a business strategy that exploits copyrighted works for profit.” In this respect, the court disagreed with the Ninth Circuit, which found in the Napster case that file sharing was commercial because “repeated and exploitative unauthorized copies were made to save the expense of purchasing authorized copies.” A&M Records, Inc. v. Napster, Inc., 239 F.3d 1004, 1015 (9thCir. 2001). In contrast, the court felt that Tenenbaum’s conduct fell “somewhere in the middle” of a spectrum of commerciality ranging from “pure, large-scale profit-seeking to uses that advance important public goals. . .”
Similarly, in its treatment of the portion of each copyrighted work infringed, the court urged that if Tenenbaum had “just sampled individual songs as a prelude to purchasing the full albums on which those songs appeared[,] [t]hat could well present a compelling argument for fair use.” Tenenbaum admitted, however, that his purpose in downloading songs was not to sample them in anticipation of later purchases, which the court ultimately acknowledged. The disappointment that results when the facts do not support a cherished theory of the case is familiar to many a litigator.
After finding that each of the four traditional fair use factors weighed against a finding of fair use, the court then addressed the creative and unusual “non-statutory factors” that Tenenbaum advanced. These included that: (1) the copyright owners assumed the risk of infringement by releasing their works in an environment where file sharing was rampant; (2) the copyright owners aggressively marketed their works while failing to protect them in any meaningful way, essentially creating an attractive nuisance; (3) Tenenbaum was forced to engage in file sharing because only entire albums, not individual songs, were available for legal purchase; (4) it is unfair for parents and universities to bear the costs of policing the file-sharing activity of children and students; and (5) the “injustice of the action” weighed in favor of fair use.
The court properly rejected each of these arguments, though it viewed some of them with a degree of approval. For instance, because the Supreme Court has suggested that the unavailability of a work for purchase through normal channels is a proper fair use consideration, the court felt that Tenenbaum was “on firmer ground” in arguing that his conduct was excused because he could only legally buy entire CDs rather than individual songs. Nonetheless, by August 2004, when Tenenbaum’s file sharing was detected by the plaintiffs, “a commercial market for digital music had fully materialized,” making the “unavailability of paid digital music  simply not relevant.”
The opinion concluded by reiterating that the court was “very, very concerned that there is a deep potential for injustice in the Copyright Act as it is currently written. It urges – no implores – Congress to amend the statute to reflect the realities of file sharing.”
This opinion – and the verdict that followed it – should strike fear into the hearts of file sharers everywhere. It is the second staggering jury verdict against an individual file sharer, following the nearly $2 million verdict in Capitol Records v. Thomas-Rasset. Nonetheless, at least one piece of anecdotal evidence suggests that file sharers are not so easily deterred: overheard in the ticket line at a movie theatre over the holidays, one youth commenting to another, “We can just download it illegally online!”