We thank Krista L. Cox, Director of Public Policy Initiatives at ARL, for the following analysis, which originally appeared on the ARL Policy Notes blog. Authors Alliance submitted an amicus brief in support of fair use in the second appeal to the Eleventh Circuit in 2017, and we continue to track the progress of the case.

The long saga of the Georgia State University (GSU) e-reserves case continues as the Court of Appeals for the Eleventh Circuit reversed the district court’s ruling which had found that the vast majority of GSU’s use of works in its e-reserves constituted a fair use. This is the second time the Eleventh Circuit has reviewed the case, and the second time it has reversed.

In 2008, publishers sued GSU for copyright infringement, arguing that the use of unlicensed excerpts of copyrighted works in the e-reserves constituted infringement. GSU defended itself, relying on the right of fair use. In the first bench trial, the district court ruled in favor of fair use for 43 of the 48 cases of alleged infringement. The Eleventh Circuit reversed and remanded the case in 2014, directing the lower court to re-examine its weight to market substitution and re-evaluate the four fair use factors holistically, rather than taking an arithmetic approach (i.e., if three fair use factors favor the use, but one disfavors it, fair use should always apply). On remand, the district court re-evaluated the four factors and found that 44 of the 48 cases constituted fair use. In her analysis, Judge Evans assigned each factor a weight: “The Court estimates the initial, approximate respective weights of the four factors as follows: 25% for factor one, 5% for factor two, 30% for factor three, and 40% for factor four.” The publishers again appealed to the Eleventh Circuit, which heard the case in 2017. (Here’s a link to ARL’s amicus brief in the second appeal.)

On October 19, 2018, the Eleventh Circuit released its 25 page opinion—more than a year after hearing oral arguments in the case—finding that the district court again erred in its evaluation of fair use. The Eleventh Circuit suggests that the district court was only mandated to re-evaluate its analysis on the second and third factors, but had instead also re-evaluated its analysis on factor four (in which the district court found in the first trial that in 31 cases, the fourth fair use factor weighed against fair use).

Additionally, the Eleventh Circuit points out that “The district court again applied a mathematical formula in its overall analysis of fair use,” which it had been instructed against. Although the district court couched the given weights as “initial” and “approximate,” the Eleventh Circuit found that the district court only adjusted these factors in four instances and di not adjust the other factors in the overall analysis. Thus, “We conclude that the district court’s quantitative rubric was an improper substitute for a qualitative consideration of each instance of copying in the light of its particular facts.” The Eleventh Circuit has remanded the case, directing the district court to use a holistic approach to fair use, and avoid any mathematical approach with respect to the four factors.

Another issue the Eleventh Circuit opinion addresses is whether the cost of purchasing licenses affects the third factor; the district court in the second trial considered the price of use on two ocassions. The Eleventh Circuit rules that price should not be taken into account when evaluating the amount and substantiality of the portion of the work used.

While the Eleventh Circuit reversed and remanded on the above issues, it affirmed the district court’s decision not to reopen the record. Publishers in 2015 filed a motion to reopen, asserting the need to introduce “Evidence of GSU’s ongoing conduct (e.g. its use of E-Reserves during the most recent academic term)” as well as evidence of the availability of digital licenses. Here, the Eleventh Circuit notes that this decision is within the discretion of the trial court.

Kevin Smith posted about the GSU case on In the Open, with an excellent summary of what the Eleventh Circuit’s opinion (as well as its last opinion) does not do, and what, as a result, the publishers have lost on:

…But the big principles that the publishers were trying to gain are all lost. There will be no sweeping injunction, nor any broad assertion that e-reserves always require a license. The library community will still have learned that non-profit educational use is favored under the first fair use factor even when that use is not transformative. The best the publisher plaintiffs can hope for is a split decision, and maybe the chance to avoid paying GSU’s costs, but the real victories, for fair use and for libraries, have already been won.

photo by 422737 |CC0

Today, the Library of Congress adopted exemptions to section 1201 of the Digital Millennium Copyright Act (DMCA) that prohibits circumvention of technical protection measures. As a part of this rulemaking process, Authors Alliance petitioned to renew the existing exemption that allows authors to bypass encryption to make fair use of film clips in nonfiction multimedia e-books offering film analysis, and we additionally advocated to modify the exemption to allow for the use of film clips in e-books for purposes other than film analysis and in fictional works.

We are pleased to report that the Library of Congress renewed the existing multimedia e-book exemptions and expanded the exemption to all nonfiction multimedia e-books (that is, the exemption is no longer limited to multimedia e-books offering film analysis). The Acting Register of Copyrights found that examples presented in our comments showed a variety of uses of short excerpts in nonfiction multimedia e-books to provide criticism or commentary beyond film analysis and also found that the brevity and transformative nature of the proposed uses favor an exemption because these uses are unlikely to substitute for the original work. The Acting Register concluded that an expansion to all nonfiction multimedia e-books is unlikely to harm, and may increase, the availability of copyrighted works and that the proposed uses will facilitate criticism, comment, teaching and/or scholarship.

The Acting Register declined to recommend to expand the exemption to cover fictional e-books, finding that the record lacked sufficient evidence demonstrating a need to expand the current exemption. We are disappointed that the Copyright Office and the Library of Congress passed on the opportunity to extend the exemption to authors of fictional multimedia e-books, such as fanfiction authors, whose transformative expression fits within the fair use doctrine.

We are grateful to the student attorneys and their supervisors at the Intellectual Property, Arts, and Technology Clinic at UC Irvine School of Law and the Samuelson-Glushko Technology Law & Policy Clinic at University of Colorado School of Law for their work supporting this exemption.

Stay tuned for additional analysis of the 1201 rulemaking for multimedia e-books from the student attorneys at UC Irvine and Colorado Law.

photo by 422737 |CC0

In December 2017, Authors Alliance submitted comments to the U.S. Copyright Office in support of a modified exemption to Section 1201 of the Digital Millennium Copyright Act (DMCA) for multimedia e-books. The proposed modification would broaden the application of the current exemption (which allows for lawful circumvention of digital rights management technologies in non-fiction works offering film analysis) to embrace fictional e-books and e-books on subjects other than film analysis. This would allow a more diverse group of creators, such as fanfiction authors, to access the clips they need from DVD and Blu-ray discs to make fair use of video clips in their e-books.

Responses to that initial comment were filed last month. Now, joined by the American Association of University Professors, Organization for Transformative Works, the Interactive Fiction Technology Foundation, and Professor Bobette Buster, we have in turn submitted reply comments as part of the U.S. Copyright Office’s seventh triennial rulemaking process for 2018. The full text of the comment is included below.

Our comment explains that creators are interested in exercising their fair use rights by engaging in criticism and commentary in their creations, and it provides evidence that creators of content have been and will continue to be harmed absent the proposed modification to the current exemption.

Our advocacy in this area is ongoing. In 2015, as part of the sixth triennial rulemaking, Authors Alliance traveled to Washington, DC to testify in hearings before the U.S. Copyright Office. We will do so again in April of this year and will keep our readers updated on that testimony, as well as the results of the seventh triennial rulemaking.

What effect does fair use have on incentives to create? For some academic authors, there is growing evidence that fair use will not diminish, and may even enhance, their incentives to create and distribute scholarly works because it promotes their goals of advancing the progress of knowledge, builds reputational capital, and increases the impact of their works.

Last year, Authors Alliance filed an amicus brief with the Eleventh Circuit in support of Georgia State University’s fair use defense in Cambridge University Press v. Becker. One issue we discussed in our brief is how our members’ experiences accord with the district court’s conclusion that academic authorial incentives to create scholarly book chapters would not be impaired by a fair use ruling. We explained that the primary motivation of academic authors to write scholarly book chapters is generally to share the knowledge and insights they have gained, and the type of reward that academic authors have generally sought and hoped to attain through writing scholarly book chapters is enhancement of their reputations.

Our brief highlighted quotes from several authors of book chapters at issue in the case who reflected on the benefit of fair use to their goals of reaching readers and contributing to academic discourse. For example, Norma Mertz, Professor of Higher Education Administration at University of Tennessee, Knoxville, wrote “I have no objection to the fair use of chapters from my books. Indeed, I find the suit to prevent use of such chapters a serious hindrance to the advancement of knowledge.” Other authors pointed to the benefits of publishing academic works. Douglas Harper, Professor Emeritus of Sociology at Duquesne University, wrote “There is reputational benefit… to doing this work. … The point of this work is to share it!”

A subsequent survey of authors conducted by Brandon Butler and David Hansen reinforced these observations about academic authors’ incentives to create. Butler and Hansen wanted to test the hypothesis that most academics expect their work to be used freely for teaching, partly because academic authors make such uses themselves. They surveyed the authors affected by Georgia State litigation—primarily academic authors whose works had been excerpted by GSU professors in support of their teaching. Their results, although based on a small sample size, suggest that academic authors expect that their works will be used for educational purposes; indeed, many of them make such uses of others’ works as well, and are not highly incentivized to write by copyright restrictions or the promise of royalties.

As Hansen and Butler write in their survey analysis, “Authors faced with a publishing contract that includes a copyright transfer or license should consider whether they trust the publisher to enforce those rights in ways consistent with academic values and expectations.” This and other issues of interest to those considering publication options will be addressed in our forthcoming guide to publication contracts. The guide will be the fourth volume in our series of educational handbooks—stay tuned for a release later this year.

In the meantime:

• Check out this model contract, which contains author-friendly terms for open access publishing of long-form scholarship,
• Stay up to date with the latest in fair use on the Authors Alliance resource page, and
• Refer to the 2018 Fair Use Week website for a collection of posts on fair use and fair dealing.

Illustration courtesy of Gary Zamchick | Used with permission

The principles of copyright law sometimes have a way of appearing in unexpected places. Recently we featured an article by Christopher Sprigman that examines assumptions about copyright as a spur to creativity by considering examples as diverse as Italian opera and Bollywood movies.

Today, as part of Fair Use Week, we are highlighting new research by NYU Clinical Teaching Fellow Amanda Levendowski that explores the ways in which copyright law can negatively influence the quality of artificial intelligence (AI), and how fair use might be part of the solution. She describes how there has been an increase in examples of AI systems reflecting or exacerbating societal bias, from racist facial recognition to sexist natural language processing.

As the computer science adage “garbage in, garbage out,” succinctly puts it, an AI system is only as good as the information provided to it. Training using biased or otherwise unsatisfactory data can result in flawed and incomplete outcomes. As Levendowski writes, “[J]ust as code and culture play significant roles in how AI agents learn about and act in the world, so too do the laws that govern them. … The rules of copyright law…privilege access to certain works over others, encouraging AI creators to use easily available, legally low-risk sources of data for teaching AI, even when those data are demonstrably biased.”

With potential statutory damages running as high $150,000 per infringed work, AI creators often to turn to easily available, legally low-risk works train AI systems, often resulting in what Levendowski calls “biased, low-friction data” (BLFD). One such example is the use of the “Enron emails”, the 1.6 million emails sent among Enron employees that are publicly available online, as a go-to dataset for training AI systems. As Levendowski puts it, “If you think that there might be significant biases embedded in emails sent among employees of [a] Texas oil-and-gas company that collapsed under federal investigation for fraud stemming from systemic, institutionalized unethical culture, you’d be right: researchers have used the Enron emails specifically to analyze gender bias and power dynamics.”

What’s more, Levendowski describes how copyright law favors incumbent AI creators who can use training data that are a byproduct of another activity (such as the messages and photos Facebook uses to train its systems) or that it can afford to purchase. This can play a determinative role in which companies can effectively compete in the marketplace.

So how can we fix AI’s implicit bias problem? In her article, Levendowski argues that if we hope to create less biased AI systems, we need to use copyrighted works as training data. Happily, copyright law has built-in tools that help to balance the interests of copyright owners against the interests of onward users and the public: One of these tools is fair use. By examining the use of copyrighted works as AI training data through the lens of fair use cases involving computational technologies, Levendowski suggests that relying on fair use to use copyrighted materials in training systems could provide a promising path forward to combat bias and make AI more inclusive and more accurate.

Read the full article on SSRN; and learn more about Levendowski and her research on her website.

Amanda Levendowski is a Clinical Teaching Fellow with the NYU Technology Law and Policy Clinic. Her clinical projects and research address how we can develop practical approaches to digital problems. Amanda previously practiced copyright, trademark, Internet, and privacy law at Kirkland & Ellis and Cooley LLP.