By Katie Fortney and Renata Ewing

Controlled digital lending has been in the news a lot lately, although libraries have been engaging in it and related practices for years. It can be difficult to keep track of all the relevant events, timelines, and activities. Since we expect that many UC HathiTrust users might appreciate a recap, as well as some notes about how this ties in to services used at UC, we've put together this summary, which provides both a review of how we got here and a snapshot of the current landscape.

The mission of libraries is to improve access to materials for their readers and researchers. In the United States, the fair use doctrine (as outlined in Title 17 section 107 of the copyright act) has long helped libraries to achieve this mission. Fair use allows copies of protected works to be made “for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research.”  In addition to these protections, section 108 carves out specific allowed activities for libraries and archives including the ability to make a copy of a copyright-protected work for the purpose of preservation, or interlibrary loan. Meanwhile, libraries make copies of and provide access to many works that aren’t protected by copyright at all, either because they’re ineligible or have fallen into the public domain.


TABLE OF CONTENTS


Fair Use and HathiTrust

Fair use is a flexible rule, designed to be able to evolve with new and developing technologies. For example, the rulings in Authors Guild v. Google and Authors Guild v. HathiTrust both held that the digitization of in-copyright books without permission from the rights holders was a fair use because it accomplished the transformative purpose of allowing full text search. This functionality breathes new life into older works, many of which are no longer in print at all, let alone available digitally. Being able to search through the tables of contents, indexes, and full text of books - rather than just their titles and authors - allows readers and researchers to find works relevant to their interests that they might not otherwise have known existed.

Fair use also allows HathiTrust to give researchers from member institutions access to data sets that include in-copyright works for non-consumptive research purposes (see Operationalizing “Non-Consumptive” Fair Use to Revolutionize Humanities Research, by Brandon Butler, HathiTrust Feb 2017). 

Neither Google nor HathiTrust allows unlimited access to the full text of in-copyright volumes, however. In either service, that kind of access is only available for public domain volumes (those published in the United States before 1929 or published outside of the US before 1899) or those volumes for which the rights holder has signed a license. In addition, as verified by the court in the HathiTrust decision, it is fair use for HathiTrust members to provide access to in-copyright works to affiliated users with print disabilities.


The Pandemic and ETAS Changed Everything

In 2020, the COVID-19 pandemic and ensuing library shutdowns changed everything. In a matter of weeks, HathiTrust heroically pulled together its Emergency Temporary Access Service (ETAS). This service was based on principles of fair use and allowed affiliates (students, faculty, and staff) of member institutions, who lost access to physical collections due to library closures, limited access to scans of in-copyright works. Member affiliates were required to log in to HathiTrust using their institution’s authentication system, allowing them access to in-copyright books their library held on its (then closed) physical shelves. ETAS users were required to “check out” the volumes, and HathiTrust maintained a one-to-one correlation between the digital and physical volumes. (For example, if the member institution had 3 copies of a title, only 3 users from that institution could check it out at the same time). There was also a one-hour time limit per checkout, though checkout could be easily renewed. For more information about ETAS at UC see:


Despite the limitations, ETAS proved to be invaluable to students and researchers who otherwise would not have been able to access volumes they needed. ETAS allowed member affiliates to continue to work on papers, theses, dissertations, articles, and books that otherwise would have sat idle during the many months of physical library closures. When the shutdowns ended and the stacks of UC campus libraries were once again open and accessible, UC’s Council of University Librarians (CoUL) requested a report to assess the impact of ETAS at UC, with the specific goal of gathering data to inform discussions and decision making about the potential use of other digital lending projects at UC. You can read the resulting report here:


UC BEARS (Controlled Digital Lending at UC Berkeley)

While the physical libraries and stacks at UC Berkeley were closed for access during the pandemic, a few library workers continued to go into the building to digitize volumes needed for student course reserves. When a book requested for a course reserve was not available for purchase via an ebook license, library staff digitized it and added it to HathiTrust where it could be accessed by students via ETAS. The system proved to be so useful and successful that Berkeley was keen to continue the practice even after the libraries were physically open and ETAS had been discontinued. For this purpose, they created the UC BEARS (UC Berkeley Electronic and Accessible Reserves System) which uses a controlled digital lending system (more on this below) for course reserves. It was inspired by the CalTech DIBS system.


Project LEND

Building on UC’s experiences with ETAS, in 2022, library leadership at UC Davis began to investigate the potential for the expanded lawful use of digitized books held by academic and research libraries. The result was UC’s Project LEND (Library Expansion of Networked Delivery), supported by the Andrew W. Mellon Foundation, which seeks to analyze all aspects of a potential digital access program (including user needs, legal frameworks, technical requirements, and collection scope). 

Project LEND is a UC-wide initiative with Rice Majors (UC Davis) as principal investigator and Eric Mitchell (UC San Diego) and Günter Waibel (CDL) as co-principal investigators. In addition to UC libraries leadership, the project steering committee includes experts from HathiTrust and the Digital Public Library of America. The goal of the project is to design an expanded service (or set of services) that could be used across UC, and that could potentially be applied by other academic libraries as well. For more about Project LEND see:


What is Controlled Digital Lending and Where Did it Come From?

Controlled digital lending is a set of practices developed by librarians and lawyers for library lending of digitized in-copyright books. The practices are designed to closely mirror physical library lending, as an exercise of fair use. Just as in physical lending, the library buys a copy of a book and then loans it out to one patron at a time. To accomplish this, the library must first digitize the book and then use digital rights management software to ensure that it is checked out and returned properly–and to prevent users from sharing or downloading the volumes. 

In 2011, Internet Archive (along with 150 library partners) launched a “Digitize and Lend” ebook lending program on its Open Library website. At the time, the lending program comprised a 80K ebook collection of mostly 20th century publications. The books were all from publishers’ backlists (not recent publications) and some of them were still in copyright. Anyone with a free Open Library account could borrow up to five ebooks at a time for up to two weeks. The program was based on a one book, one reader model. That same year, Michelle Wu, an attorney and librarian and the director of the Georgetown Law Library, published “Building a Collaborative Digital Collection: A Necessary Evolution in Libraries.” In this article she presented what she called “a vision of a collaborative, digital academic law library” that sought to identify and respond to the difficult issues (including copyright) that surrounded collaborative library digitization projects. Her paper was “intended to be a call for collective action—to stop discussing the law library of the future and to start building it.” She called her proposed project TALLO (Taking Academic Law Libraries Online) and proposed digitizing volumes and permitting the circulation of the exact number of copies purchased to stay within, what she argued, to be the bounds permitted by copyright law. 

Wu, along with Internet Archive Policy Council Lila Bailey, was one of the co-authors of the Position Statement On Controlled Digital Lending by Libraries in 2017. The statement offered a “good faith interpretation of U.S. copyright law for American libraries considering how to perform traditional lending functions using digital technology.” It outlined the two main areas of copyright law that support controlled digital lending: a) first sale and the common law exhaustion principle and b) fair use, including an analysis of the four fair use factors in relation to controlled digital lending. Two of the other co-authors of the statement, David R. Hansen and Kyle K. Courtney, subsequently wrote A White Paper on Controlled Digital Lending of Library Books. The White Paper more fully explained the legal rationale underlying controlled digital lending, and then discussed the risk factors and practical considerations relevant to the design of a controlled digital lending system.

In subsequent years, other individuals, libraries, and organizations have built on the idea of controlled digital lending and how it might be used. For example:


National Emergency Library and Lawsuit

The Internet Archive and its library partners continued to build the Open Library’s ebook lending collection following its launch in 2011. In response to the COVID-19 pandemic shutdown of libraries, Internet Archive opened its National Emergency Library on March 24th, 2020, “to address our unprecedented global and immediate need for access to reading and research materials.” The Open Library’s collection of ebooks was made available, temporarily, without the existing lending restrictions. Anyone could check out any book with no wait list, even if this resulted in the number of checkouts exceeding the number of physical volumes owned by Internet Archive and its partners. The move resulted in both positive and negative coverage in the press. For an example of each, see below:


On June 1, approximately two months after the announcement of the National Emergency Library, four publishers (Hachette, HarperCollins, Wiley, and Penguin Random House) filed a lawsuit against Internet Archive. On June 16, the National Emergency Library shut down, two weeks before its planned closure of June 30, a date chosen to coincide with the US academic calendar. The lawsuit did not single out the National Emergency Library service; it claimed that the entire Open Library practice of lending digitized, in-copyright works without publisher permission was copyright infringement. For more coverage of the publisher lawsuit from before there was a court opinion, see:


The case was filed in federal court in the Southern District of New York. The publishers included a list of 127 books that they made available as licensed ebooks, which Internet Archive had loaned through Open Library and the National Emergency Library. The plaintiffs (publishers) and the defendant (Internet Archive) both made motions for summary judgment, asserting that the law was clearly enough in their favor that the judge could rule in their favor without a trial. Other groups and individuals who were not parties to the case filed amicus briefs supporting one side or the other. All these filings, and more, are publicly available from the district court case docket on the CourtListener website. 

On March 24, 2023, Judge Koeltl granted the publishers’ motion for summary judgment and denied Internet Archive’s in a 47-page opinion. The opinion ordered the parties to work out some of the details regarding implementation of the judgment. These details were finalized in August 2023; Internet Archive subsequently appealed. For more about the district court decision, see:


The case is currently on appeal to the Court of Appeals for the Second Circuit. Once again, many amicus briefs have been filed. All of them can be found in the appellate court case docket on the CourtListener Website. Some of the amicus briefs have been summarized by commentators:


Oral arguments in the appeal were heard on June 28, 2024 (a recording is available from the Second Circuit's Oral Argument Recordings page; search for "Hachette"). A decision will follow, but it’s difficult to guess how soon. Whoever prevails on appeal, the Second Circuit’s opinion in this case is sure to have plenty of interesting things to say about fair use, and the library community will read it closely. Issues of fair use are always fact-specific, however, and libraries are always faced with new use cases. For better or worse, the opinion is unlikely to answer all questions about current and future library uses.