Tag: fair use

ReDigi, Fair Use, and Libraries

Photo by author.

In my last post, I opined that the fair use interests of librarians and educators are not necessarily aligned with for-profit business ventures seeking to exploit creative works in ways that can harm authors.  For instance, in the case of Capitol Records v ReDigi, now on appeal at the Second Circuit,  Jonathan Band filed an amicus brief on behalf of the American Library Association, the Association of College and Research Libraries, the Association of Research Libraries, and the Internet Archive in favor of a reversal of the district court opinion, which held that ReDigi is infringing.

To the casual observer, ReDigi is a service that acts as a middle-man to broker the sale of “used” MP3 files.  The idea is that, when you decide you’ve listened to your iTunes-purchased copy of Meat Loaf’s “Bat Out of Hell”* for the last time and think somebody else might want it,  you can sell the MP3 files via ReDigi as “used” items for a portion of the original price, with a cut going to ReDigi.  From a legal standpoint, however, ReDigi is a somewhat complex, computer-to-computer model hoping to avail itself of a copyright limitation written for a purely physical world.

To the best of my knowledge, the ReDigi model has changed a couple of times, but the way the system works is that the MP3 files being offered for sale are removed from the seller’s devices and copied onto the ReDigi server from which the seller alone may still listen to the tracks until they are purchased by someone else.  Upon sale, the files are removed from the server and copied again onto a device belonging to the purchaser. Naturally, this sounds attractive to many consumers, and ReDigi seems to have made an effort to ensure, to the best of its ability, that only one owner at a time has the files.  But that doesn’t necessarily make the service legal, or even a good idea in the market.

The district court held that ReDigi infringes the plaintiff’s rights of reproduction and distribution. ReDigi’s defenses that it is non-infringing were based on the first sale doctrine and fair use; and it is with regard to the fair use defense that the libraries apparently see a common interest.

Band’s amicus brief argues that a finding of fair use in ReDigi would set a precedent that furthers the interest of libraries to develop new models for digital-age lending and other services. And this is where I personally see a necessary distinction between the interests of libraries—which sometimes receive specific carve-outs in copyright law—and avoiding the particular harms caused by a ReDigi model, which could very easily be an Amazon-scale service overnight.

Digital First Sale Creates an Artificially Deflated Market

The first sale doctrine (Section 109(a) of the copyright law) is the reason you are allowed to sell your individual copies of books, CDs, DVDs, etc. as used items without infringing the copyright interest of the authors.  Its predicate is a case from 1908, long before anyone could possibly imagine “owning” entertainment media in the form of data that is not exclusively bound by a physical object like paper or a disk.   ReDigi argued that the first sale doctrine makes its business non-infringing, but the court disagreed on statutory grounds—namely that the language “plainly applies to the lawful owner’s ‘particular’ phonorecord, a phonorecord that by definition cannot be uploaded and sold on ReDigi’s website.”

As a technical matter, ReDigi has to make a copy of the files from the “seller’s” computer to ReDigi’s server. That copy is already an infringement of the rights holders’ exclusive right of reproduction, and that copying is not protected by first sale doctrine.  But even if one wishes to argue that this is a mere technicality, there is still ample reason not to apply first sale doctrine in a purely digital market.

The District Court stated that if ReDigi seeks amendment  to the Copyright Act to satisfy its interpretation of first sale, that this is job of Congress and not the courts. On that subject, Congress has reviewed first sale doctrine as part of its comprehensive review of the Copyright Act. (In fact, I sat on a panel in 2014 hosted by the USPTO discussing this very issue.) And the Department of Commerce examined the subject of digital first sale and concluded that the principle did not apply to digital transmissions and that there is no evidence warranting any change to existing law.

I advocate against digital first sale for the simple reason that digital files cannot be “used” according to any rational definition of that term.  Since 1908, the distinction between the used, secondary market for these goods and the primary market has been clear, with the former posing no threat to the latter.  Even in cases in which used items appreciate in value (e.g. rare books), I would argue this will only ever apply to physical goods and not to raw data that may be endlessly copied from device to device. (A signed, first edition Yeats is worth a couple grand; an eBook containing the same written material is free.)

Because a “used” digital file is as pristine as a “new” digital file, a service operating at scale like a ReDigi (or an Amazon) would simply create a new primary (or alternative) market that artificially degrades pricing for digital-only works.  Regardless of the fact that some consumers feel that any price above zero is a “rip off,” the reality is that we pay comparatively low prices for the media we store today, and even lower prices for media we access through subscription or ad-funded models.  If anything, Congress should reaffirm that first sale applies solely to physically-bound, legally acquired, copies of works.

ReDigi’s Fair Use Argument and the Libraries’ Interests

Apparently unconcerned with the potential harms of a ReDigi model in the consumer market, the above-named library organization would like the appeals court to find that ReDigi’s service is a fair use under the first prong of the fair use test:  1) The purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit purposes.

Initially, ReDigi sought to defend its position as a cloud service—that users have a fair use right to store their files in the cloud; and Capitol did not dispute this part of the argument.  But the fact that this storage was “incident to sale” of the works meant that ReDigi’s service falls “outside the ambit of fair use.”  The district court agreed.  Moreover, the court held that ReDigi’s use was not “transformative” because its service does not add anything to the works and that the commercial nature of the use tilted away from a finding of fair use.  The other three prongs of the fair use test do not favor ReDigi, particularly the fourth, given that their use can clearly cause market harm to the rights holders.

In his brief, Jonathan Band takes a curious position that the lower court failed to understand how ReDigi is protected by first sale doctrine and that a reversal of this reasoning would support a finding of fair use.  He writes …

“The district court overlooked the obvious fact that the use ReDigi sought to enable was analogous to one permitted by a central feature of the Copyright Act: the first sale right, codified at 17 U.S.C. § 109(a). The district court found that ReDigi’s uses did not fall within the scope of section 109(a) because the first sale doctrine is a limitation on the distribution right, not the reproduction right, and ReDigi’s service involves reproduction. However, the similarity between ReDigi’s service and conduct permitted under section 109(a) should have weighed in ReDigi’s favor under the first fair use factor.”

Going back to what first sale allows—namely that you are free to sell that “Bat Out of Hell” CD at a yard sale—fair use doesn’t enter the picture at all. Your right to sell your particular copy of a book or album is not a use of a copyrighted work that in any way implicates a fair use defense.  So, even if these petitioners could persuade the court that it is sufficient that ReDigi is “analogous” to the application of the first sale principle, it is hard to imagine how this favors a finding of fair use, which is supposedly the real interest among these librarians.

My read is that the argument Band presents seems over-broad—a hope that a finding of fair use in ReDigi will open a wider door for libraries to develop new means to fulfill their missions in a digital world.  In this sense, Band appeals to the Google Books case, in which Google Books prevailed on fair use analysis to the benefit of library interests. But without writing another thousand words, suffice to say that Google Books is nothing like ReDigi.  Google Books does not engage in commercial transactions, it does not make whole works available, and it provides services that are either “transformative” (e.g. full-text search) or fair use prima-facie (making works available to the print-disabled).

A finding of fair use in ReDigi may generically support the interests of libraries (or it may not), but it seems a high price to pay to allow a harmful, for-profit company to vitiate fundamental copyright protections. As implied, this paves the way for a massive organization like Amazon to decimate what remains of the primary market for sales of creative works.  I believe these matters should be viewed separately, with libraries and universities pursuing their own interests in context to copyright law rather than riding the coattails of predatory and opportunistic business ventures.

*Nothing personal, Meat Loaf. First album that came to mind.

Masnick Makes a Hash of Fair Use & Censorship

Photo by Pond5
Photo by Pond5

In an effort to conflate president-elect Trump’s rhetoric on censoring the press with copyright protection, Mike Masnick at Techdirt accuses the News Media Alliance of seeking to “whittle down” fair use. He further says this will only leave journalists vulnerable to the kind of censorship Trump has threatened by amending libel laws.  There are too many holes in Masnick’s post to address efficiently, so I’ll stick with the main point about fair use doctrine. The Newspaper Association writes the following:

“Fair use” should be reoriented toward its original meaning. Under current copyright law, a person that does not own a copyright may still use a copyrighted work if it is consistent with the “fair use” factors, which assess: (1) the purpose and character of the use, (2) the nature of the copyrighted work, (3) the amount and substantiality of the portion taken, and (4) the effect upon the potential market. The courts, unfortunately, have dramatically weakened this test by finding a fair use any time a new use could be seen as “transformative.” This test has undermined the integrity of the long-established fair use factors. As part of any Copyright Act rewrite, we support refocusing the fair-use test on its original purpose to prevent courts from undermining the Constitution’s encouragement of compensation to entities that generate creativity and productivity.”

For starters, this statement isn’t asking anyone to “whittle down” fair use. Instead, the News Media Alliance is simply asserting what many copyright experts and rights holders have observed, which is that the “transformative” standard is in fact a relatively new and often-vague principle that has become something of a vestigial fifth factor not codified in the 1976 Copyright Act.  In fact, “transformativeness” began as a measurement of creative transformation in the landmark case Campbell v Acuff-Rose but has since been applied in broad contexts in which uses are “transformative” of something other than the original work to create a new expression.  

So, “transformativeness” can exceed the original free-speech motivations for codifying fair use into the federal law in the first place.  And that in itself is not inherently bad; we want law to be elastic to a certain extent, otherwise copyright itself could not have adapted to changing market and technological conditions. 

Having said that, however, the “transformative” standard has come dangerously close to asserting that simply using a work in a new context—like posting it on social media—is “transformative” enough to make the use fair.  So, the Alliance is not attacking fair use doctrine at all, as Masnick asserts, but is rather seeking to mitigate what many rights holders view as an irrational expansion of the doctrine until it ceases to be an exception at all.  

The part where Masnick accuses the Alliance of playing into Trump’s censorship hands is just a malarky cocktail well spun.  He writes the following:

“While [Trump] was specifically talking about libel laws, as we’ve seen over and over again, copyright is an amazing tool for censorship as well. In fact, the Supreme Court itself has noted that fair use is the necessary “safety valve” on copyright’s free speech stifling powers. So for newspapers to basically gift wrap to Trump a way in which he can pull back a tool that protects their free speech — just as he’s been promising to attack their free speech — is ludicrous.”

Masnick is mashing up unrelated topics to argue the interests of OSPs like Google and taking the opportunity to use the words copyright and censorship in the same sentence. As a general statement, it is true that fair use is a free-speech-based exception to copyright, but most speech-related, or press-related, uses almost always relate to other forms of expression, including journalism, and they rarely implicate the “transformative” standard being referred to by the News Media Alliance. 

For instance, I noted in a past post that a FOX Network initially sought to argue that its use of another news agency’s photograph was “transformative” simply because it was posted on their Facebook feed.  That argument didn’t get very far, but it’s the kind of argument rights holders are nervous about arriving in the courts; and it has nothing at all to do with legitimate concerns about a president threatening to use libel laws to silence the press. For another perspective on how the “transformativeness” standard can come very close to effectively obliterating copyright, see this post about TVEyes v FOX News.  

As usual, the internet industry and its advocates behave as though their platforms, which make unlicensed uses of all manner of works, are synonymous with free speech or freedom of the press.  From that premise, they argue that a desire to maintain boundaries and contours around the fair use doctrine is synonymous with trying to kill the doctrine outright.  That is ludicrous.

TVEyes v FOX a Huge Deal for Copyright

One of the most important copyright cases in the country landed at the Second Circuit Court of Appeals in December of 2015. Appellants TVEyes and Fox News Network filed opening briefs, each seeking new rulings after the Court for the Southern District of New York held a split decision last year that satisfied neither party.  In July of 2013, Fox sued TVEyes—a B2B, subscription-based, news-monitoring service that records, stores, and indexes 1,400 channels—for copyright infringement of its programs.  TVEyes provides several features for its subscribers, some of which were held to be fair use by the district court while others were not. Most notably, its Content Delivery Features were held to be “necessarily infringing,” and thus the company was enjoined from further offering these services to its customers.

TVEyes is appealing on the grounds that the district court erred in not finding its whole business model, including the Content Delivery Features, a fair use of Fox’s copyrighted news broadcasts and that the court “abused its discretion” in enjoining the company from providing those services. Fox is appealing on the grounds that TVEyes is grossly misrepresenting its business model, the character of its customer base, and especially the nature of the Content Delivery Features; and it is seeking to have the entire district court ruling overturned. This case is important because, if the evidence presented in the Fox brief is an accurate portrayal of TVEyes, and if the business as it stands were held to be a fair use, that outcome could be devastating to the existence of copyright law itself.

It would be difficult to adequately summarize all of the issues raised by the combined 200 pages submitted by the appellants, but the significance of this case seems to rest on two major factors.  The first is the true nature of TVEyes’s business model; and the second is the manner in which TVEyes is asking the court to interpret both the first and fourth factors for consideration of fair use.

In simple terms, TVEyes portrays its service—so stated in its brief—as a “research only” tool akin to Google Books, and their argument relies heavily on this precedent ruling to assert that its entire service—including the Content Delivery Features—makes fair use of Fox’s copyrighted works.  But based on the evidence presented in the Fox brief, TVEyes provides services that not only excede the purposes of research, but which directly substitute existing, legal services for achieving the same commercial and non-commercial ends achieved by TVEyes.

Subscribers to TVEyes, who pay $500/month, include both private and public entities that can search any of the monitored networks for broadcasts of interest.  The TVEyes brief states that the company’s User Agreement requires subscribers to make exclsively internal use for research purposes only. In support of this claim, the brief cites examples that sound both niche and internally analytical, like “The Army uses TVEyes to track media coverage of military operations to ensure national security and safety of troops.” TVEyes’s testimony paints a picture of limited use and unprecedented capability in order to liken the service to that of Google Books.

The evidence presented in the Fox brief, however, offers a very different view of TVEyes—one that cannot be compared to Google Books for one simple reason that Google does not make whole, copyrighted books available to users, while TVEyes apparently does exacly this and quite a bit more.  According to Fox’s testimony, TVEyes provides alternative, infringing access to its subscribers who may a) view 24/7 live streaming of broadcasts; and b) download entire programs in sequential, ten-minute segments.  These video clips are also made available in high-definition without watermarks or copyright statements, and these facts alone are highly suspect since there is no need for either a news-monitoring service or its customers to cover the cost of high-def streaming for “research-only” purposes like those described in the TVEyes brief.

The high-def feature can, however, be explained by the evidence cited by Fox, which states that the majority of TVEyes’s customers (including the Army) appear to be PR firms or PR departments, to whom TVEyes has clearly advertised their service as a tool for storing and distributing unlimited TV content via email, social media, or other platforms.  Citing several examples from research, marketing materials, the Terms of Service, and TVEyes executive testimony, Fox presents a compelling argument that TVEyes is far from an internal-use-only research tool as claimed. Instead, Fox asserts that evidence proves TVEyes to be an “all-you-can-eat” subscription service that directly substitues live TV broadcast, the network’s own online offerings, and the network’s licensing regimes for PR and other uses identical to those being provided by TVEyes.

A classic example of a TVEyes use by a typical subscriber would be when an executive is interviewed by Fox Business Network, and her company wants to exploit the interview for PR purposes and/or edit the segment into an internal video, intranet, or closed-circuit news feed.  Fox (or any network) will normally license clips for these purposes; or if the use is a social media share, the Fox web portal already makes customized embeds possible according to Fox’s testimony.  TVEyes’s Content Delivery Features appear to circumvent both of these revenue streams that rightly belong to the network; and as the Fox brief notes with specific regard to the loss of online traffic through embeds, this kind of expropriation serves to hamper the development of digital strategies for all traditional news and entertainment producers.

The Fair Use Arguments

The most important doctrinal issue in this case is the manner in which TVEyes appears to be presenting its fair use defenses; and the two arguments of greatest consequence are the assertion of “transformativeness” under the first factor and the claims pertaining to market value harm done to Fox under the fourth factor.  Both arguments ask the court to read the statutes in entirely novel ways—interpretations that would come extremely close to nullifying copyright protection of any kind for anyone.

TVEyes’s claim of “transformativeness” relies on the manner in which some of its customers may use its suite of services. But as attorneys for Fox point out—and they specifically cite Google Books here—a defendant bears the burden of proving that its use is intrinsically “transformative” (i.e. novel) rather than argue that certain users of its service might make individual, “transformative” uses of works the defendant has made available to them. I know that’s a mouthful, but TVEyes is essentially arguing that if some of its subscribers make “transformative” uses of Fox news segments, this reflexively makes TVEyes a “transformative” service.

Fox appropriately argues that TVEyes’s assertion here “inverts the first factor” because its business merely substitutes existing, legally available (and well-established) services for both non-commerical and commercial TV-monitoring purposes.  Thus, the public interest is already served in this regard (and “the public” cannot afford TVEyes by the way) and that TVEyes is, therefore, not intrinsically “transformative.”  By contrast, Google Books was held to be intrinsically “transformative” no matter how people use it and because it does not provide a substitute for legal access to whole books but only makes “snippets” from books available.

Under the fourth factor, TVEyes states that Fox has not yet produced any evidence that it has sustained any market harm due to TVEyes operations.  Again, the Fox brief points out that this is a misinterpretation of the fourth factor analysis, which considers potential market harm and not market harm done up to the date of litigation. In addition to the fact that Fox does present examples of direct loss (e.g. 140,000 social media shares for PR purposes that should either be licensed or embedded from Fox’s site), the Fox brief correctly points out that TVEyes is asking the court to consider the fourth factor in a novel and unintended manner.  If an infringement suit is brought shortly after a use is made, and the claim is then countered with a fair use defense, potential market harm is the only rational consideration that can possibly be made at the time.  A court would not say to a plaintiff, “Let’s give the defendant a year or two and see if his use has an adverse effect on your sales.”  That’s not how the fourth factor fair use analysis is applied.

Even in this long post, I’ve only scratched the surface of the many details in this particular case.  But assuming the evidence provided by Fox is accurate, the manner in which TVEyes appears to be asking the courts to reinterpret fair use doctrine seems to be the existential crux of this case for the future of copyright.  There appears to be sufficient evidence that TVEyes is not merely using whole works as a database to create an index for research—as is the case with Google Books—but is making whole TV programs availalbe for mass, unlimited distribution.  As such, a finding of fair use for TVEyes appears to this layman to be tantamount to saying that copyright is no longer relevant.