Authors Alliance, joined by the Library Copyright Alliance and the American Association of University Professors, is petitioning the Copyright Office for a new three-year exemption to the Digital Millennium Copyright Act (“DMCA”) as part of the Copyright Office’s eighth triennial rulemaking process. If granted, our proposed exemption would allow researchers to bypass technical protection measures (“TPMs”) in order to conduct text and data mining research on literary works that are distributed electronically and motion pictures. Last week, Authors Alliance responded to questions posed by the Copyright Office as it considers the merits of the proposed exemption following last month’s public hearing on the exemption.

Text and data mining (“TDM”) refers to automated analytical techniques aimed at analyzing digital text and data in order to generate information that reveals patterns, trends, and correlations in that text or data. TDM has great potential to enable groundbreaking research and contribute to the commons of knowledge. As a highly transformative use of copyrighted works done for purposes of research and scholarship, TDM fits firmly within the ambit of fair use. But TDM researchers are currently hindered by Section 1201 of the DMCA, which prohibits the circumvention of TPMs used by copyright owners to control access to their works. Section 1201 makes TDM research on texts and films time consuming and inefficient—and in some cases, impossible—working against the promotion of the progress of knowledge and the useful arts that copyright law has been designed to incentivize.

Following last month’s hearing, the Copyright Office asked proponents and opponents of the proposed exemption to: 1) describe minimum security measures eligible institutions should be required to use to secure corpora of literary works or motion pictures used for TDM research and 2) share views on potential regulatory language that would limit a researcher’s ability to view literary works or motion pictures included in corpora. In addition, opponents were given the opportunity to respond to changes proponents proposed in our reply comment to address opponents’ concerns.

Security Standards and Controls

With respect to security measures, our response describes the flexible process that information security and data management professionals at research institutions use to select and apply security controls to research data. This approach tracks the processes laid out in international standards and federal agency procedures. We explain why these risk assessment frameworks are superior to a globally applied fixed list of minimum security requirements and how they are consistent with the Office’s approach to information security in previous exemptions.

Our letter provides examples of common and effective security controls used in many research settings, including user authentication, use of encryption, event logging, and maintaining physical security of the resources housing the data. We recommend that the Office should identify these controls as examples of reasonable security measures, while leaving room for information security departments and researchers to fine-tune the precise security controls used to the specifics of the research corpus and the information system in which it is housed.

Prohibiting Researchers from Viewing Text and Images

With respect to the extent to which regulatory language should limit a researcher’s ability to view literary works or motion pictures included in corpora, we clarify that while researchers do not need this exemption for the purpose of viewing the full text or images of the works that they or their institutions have already obtained lawfully, researchers must be able to verify their research methods and research results. This verification requires that researchers have some ability to view corpus text or images. That ability is consistent with the research environments of both HathiTrust Data Capsules and Google Book Search, and it is consistent with fair use precedent.

Our letter explains why researchers need to view enough of a corpus to verify their methods and their findings. By way of example, if an algorithm tells the researcher that frame #133292 of a corpus copy has a high probability of being a scene of violence, and that frame corresponds to a scene in the film Pulp Fiction, the researcher would not watch a copy of the DVD or digital download in its original format to verify that finding. But at some point, either the researcher or peer reviewer may need to locate and examine frame #133292, a designation that exists only in the corpus copy, to verify the algorithmic finding. Our response explains that a blanket prohibition on viewing text or images would comprehensively undermine TDM research relying on the exemption and would provide little added value or protection given the other restrictions in the proposed exemption. For this reason, we recommend that—although we do not believe an express viewability limitation is warranted—should the Office choose to include one, it should use the model of the HathiTrust Research Center’s Non-Consumptive Use Policy rather than an outright ban on viewability.

* * *

The Librarian of Congress is expected to issue a final decision on the proposed exemption in October 2021. We will keep our members and readers apprised of any updates on our proposed exemption as the process moves forward. We’re grateful to law students and faculty from the Samuelson Law, Technology & Public Policy Clinic at UC Berkeley Law School for their work supporting our petition for this new exemption.

Authors Alliance, together with Professor Ariel Katz and represented by Sana Halwani of Lenczner Slaght, participated in oral arguments before the Supreme Court of Canada in Access Copyright v. York University. Our arguments furthered the views we shared in our factum submitted last month.

The case involves a claim by Access Copyright, a Canadian copyright collective, which seeks to have York University comply with an interim tariff approved by the Copyright Board of Canada for works in its collection. In response, York University brought a counterclaim seeking a declaration that its guidelines for copying materials for education purposes constituted “fair dealing” under the Copyright Act of Canada (fair dealing is the Canadian analogue to fair use). The case raises the question of whether copyright collectives can force users to license content from them, even if the users prefer to comply with their copyright obligations in other ways.

At oral argument, we began by explaining to the court that Access Copyright does not represent the interests of all authors. Authors Alliance represents authors whose primary concern is their works having the greatest possible impact by reaching the largest possible audience. Unfortunately, the flawed approach to fair dealing taken by the courts harms these interests, and undermines our members’ efforts to support education and informed public discourse, by creating a chilling effect on the dissemination of copyrighted works. Our members’ dissemination goals are advanced by a robust approach to fair dealing.

We further explained how the fair dealing factors in this case were incorrectly dealt with because they were not anchored to specific instances of alleged infringement. The abstract nature of that inquiry was a result of the lower courts’ willingness to make a determination of infringement outside of an infringement action without the proper parties and necessary evidence. The trial court concluded that there were reproductions that entitled Access Copyright to royalties—that there was infringement—without identifying any particular reproduction that was not fair dealing.

One example of why this approach is problematic is illustrated by the way the lower courts handled the “effect of the dealing” factor. The effect of the dealing factor is intended to consider the market impact of the defendant’s actions with respect to the plaintiff’s work on the sale of or royalties from that particular work. But instead, the lower courts looked to the effect on the market generally. This general market approach untethers the analysis from the economic interests of the specific authors of the works at issue, instead bringing in irrelevant evidence of copying by other institutions and of the impact of the copying on the sales of other works.

This was a mistake and the lower courts should have addressed only whether the reproduced work adversely affects or is likely to compete with the original work, not with the market generally. We asked the court to find that it was an error of law to consider this factor in aggregate and at a market-wide level, and to reaffirm the correct approach to the effect of the dealing factor is an investigation into the effect of a specific dealing on a specific work.

Authors Alliance is grateful to Sana Halwani for skillfully representing our interests at oral argument, and to the entire Lenczner Slaght team, including Paul-Erik Veel, Jacqueline Chan, and Anna Hucman, for pro bono assistance with this intervention. We will keep readers updated on the outcome of the case.

Authors Alliance, together with Professor Ariel Katz, has submitted a factum to support our intervention in Access Copyright v. York University, currently on appeal before the Supreme Court of Canada.

The case involves a claim by Access Copyright, a Canadian copyright collective, which seeks to have York University comply with an interim tariff approved by the Copyright Board of Canada for works in its collection. In response, York University brought a counterclaim seeking a declaration that its guidelines for copying materials for education purposes constituted “fair dealing” under the Copyright Act of Canada. The case raises the question of whether copyright collectives can force users to license content from them, even if the users prefer to comply with their copyright obligations in other ways.

As our factum explains, in the absence of specific allegations of copyright infringement from copyright owners, the lower courts should not have dealt with the issues of infringement and fair dealing. Because the lower courts did so without the proper plaintiffs, the result was a misguided approach to fair dealing that undermines users’ rights and the interests of many authors. Our factum also explains that even when their works are published under “all rights reserved” models, many of our members believe that their interests are best served with a robust application of fair use and fair dealing that does not unduly interfere with their dissemination goals, particularly in educational contexts.

On the issue of whether the approved tariffs are mandatory vis-à-vis users, our factum supports the Federal Court of Appeal’s finding that the approved tariffs bind copyright collectives but cannot be imposed on users as mandatory tariffs. We highlight some of the incoherent outcomes that would follow from the mandatory tariff theory, including the further marginalization of authors who are not a part of the collective’s repertoire.

To read our factum in full, click here.

The Supreme Court of Canada will hear oral arguments in the case on May 21, and Authors Alliance has been granted permission to present up to five minutes of oral arguments at the hearing. Authors Alliance is grateful to Lenczner Slaght attorneys Sana Halwani, Paul-Erik Veel, and Jacqueline Chan, as well as law clerk Anna Hucman, for pro bono assistance with this intervention.

Yesterday, Authors Alliance submitted a comment to the Copyright Office in response to a Notice of Inquiry regarding developing regulations to govern the copyright small claims procedure under the newly enacted CASE Act. In the past, we have spoken out in favor of a sensible copyright small claims process, but cautioned that the CASE Act could invite abuse and pose a high likelihood of harm to authors as both claimants and respondents. Now, Authors Alliance welcomes the opportunity to provide our feedback to the Copyright Office so it can work to ensure that the Copyright Claims Board (“CCB”), which will hear copyright small claims, is an efficient, effective, and respected forum, and moreover that it serves the individual authors and creators it is intended to benefit. We summarize our input below, and invite you to read our full comment for more detail.

Notice Sending

One of the areas in which the Copyright Office requested input is the contents of the notices that will be sent to respondents when a claimant makes a claim against them before the CCB. The CASE Act mandates that these respondents be given an opportunity to opt-out of the proceedings and instead have the claim proceed in federal court, and requires the CCB or an entity acting on its behalf to send respondents two separate notices notifying the respondent of the claim against them. Regarding the contents of these notices, Authors Alliance advocated for clear, comprehensive, and informative notices which will convey to the recipients the nature of the CCB proceedings and the consequences of failing to opt out. We also requested that the CCB include information on these notices about why a respondent might want to opt in or opt out. If a respondent fails to opt out or appear before the CCB, they may be subject to a default judgment that is subject to limited review in federal court. It is our hope that if the Copyright Office implements our suggestions, notices will not be ignored, which could leave unwary respondents on the hook for damages. 

Opt-Out

The Copyright Office also sought guidance on the opt-out procedure for respondents who want to opt out of the CCB proceedings. Authors Alliance strongly urged the Office to make opting out as simple as possible for respondents of different levels of technical and legal sophistication. We suggested the Office allow respondents to opt out using a variety of methods: email, online form, over the phone, or by standard mail. We also encouraged the Office to develop a publicly available list of entities that intend to opt out in order to make the forum more efficient for claimants, who can check this list to see if the party they are pursuing a claim against indicates an intent to affirmatively opt out before filing a claim. Finally, we provided feedback on a special opt-out provision for libraries and archives, advocating for a robust opt-out provision that would allow libraries and archives to avoid the cost of defending excessive claims and spend their precious resources elsewhere to further the public good.

Limiting Cases

The Copyright Office also asked for input on whether and how to limit the number of cases that a given party can bring before the CCB over a calendar year. While Authors Alliance did not propose a specific threshold (which would be difficult if not impossible to do without knowing the what the CCB’s caseload will look like), we did commend the Office for its attention to this matter, and suggested that it impose meaningful limits on the number of cases that can be brought by a given party, with the overall goal of deterring unscrupulous actors while keeping the forum open and accessible to those who most need it—individual creators and authors seeking to enforce or defend their rights. We also suggested that the CCB implement regulations to deter “repeat players” from bringing repeated and ill-founded claims.

Guidelines on Unsuitability for the Forum and Award of Statutory Damages

Authors Alliance also suggested that the Copyright Office develop sets of guidelines for the CCB to use when determining whether a particular claim is appropriate for the forum and guidelines for the award of statutory damages. The CASE Act requires that the CCB dismiss claims that are not suitable for CCB adjudication, but does not provide much in the way of guidance as to how to determine whether the CCB is a suitable forum for a given claim. We suggested that the Office set guidelines to help the CCB determine whether a case is suitable for the proceedings, encouraging the CCB to dismiss complicated, fact-specific claims, and hear only straightforward infringement claims. Complicated, fact-specific issues like fair use are not appropriate for this streamlined procedure, and guidelines to this effect would go a long way to making the forum efficient and accessible. Regarding statutory damages, we suggested that the Copyright Office issue guidelines for the CCB to use when deciding whether to award statutory damages. In general, damages awards be proportional to the actual harm from the alleged infringement, rather than the maximum allowable damages under the statute ($15,000 per work and no more than $30,000 overall), which is often grossly disproportionate to lost licensing revenue the claimant would have received if the alleged infringer had obtained a license to use the work. We also suggested that the CCB should be particularly mindful to avoid statutory damages in cases of noncommercial uses.

Authors Alliance, joined by the Library Copyright Alliance and the American Association of University Professors, is petitioning the Copyright Office for a new three-year exemption to the Digital Millennium Copyright Act (“DMCA”) as part of the Copyright Office’s eighth triennial rulemaking process. If granted, our proposed exemption would allow researchers to bypass technical protection measures (“TPMs”) in order to conduct text and data mining research on literary works that are distributed electronically and motion pictures. Yesterday, Authors Alliance participated in public hearings hosted by the Copyright Office to consider the merits of the proposed exemption.

Text and data mining (“TDM”) refers to automated analytical techniques aimed at analyzing digital text and data in order to generate information that reveals patterns, trends, and correlations in that text or data. TDM has great potential to enable groundbreaking research and contribute to the commons of knowledge. As a highly transformative use of copyrighted works done for purposes of research and scholarship, TDM fits firmly within the ambit of fair use.

But TDM researchers are currently hindered by Section 1201 of the DMCA, which prohibits the circumvention of TPMs used by copyright owners to control access to their works. Section 1201 makes TDM research on texts and films time consuming and inefficient—and in some cases, impossible—working against the promotion of the progress of knowledge and the useful arts that copyright law has been designed to incentivize.

At yesterday’s hearing, the clinical team from the Samuelson Law, Technology & Public Policy Clinic at UC Berkeley Law School representing Authors Alliance testified about the details of the exemption and its immense value for TDM researchers. The team explained how section 1201 prevents those researchers from creating the corpora of works they need to discover new insights from text and data mining, interfering with their ability to generate new copyrighted works that add to our cultural understanding and advance human knowledge.

Specifically, clinic students, Ziyad Alghamdi, Tait Anderson, and Erin Moore, and clinical supervisor, Professor Erik Stallman, shared how section 1201’s prohibitions chill new research and hinder the progress of knowledge in at least three ways: 1) forcing researchers to limit datasets in a way that makes their findings less illuminating than they would otherwise be, 2) causing researchers to artificially constrain research to public domain texts, and 3) leading researchers to abandon potential TDM projects altogether.

Opponents of the exemption testifying in the hearing—representing publishers, the software industry, and content licensing organizations— raised concerns about whether TDM was fair use under copyright law, whether the proposed security measures for the TDM corpora were sufficient to allay their security concerns, and whether alternatives like pre-assembled TDM corpora would be adequate for TDM researchers.

Regarding fair use, Erin Moore testified that relevant case law firmly establishes TDM as a fair use, and that the fact that the use could have been officially licensed by the copyright holder does not mean the use is not a fair one. Moore also emphasized the noncommercial and educational nature of the uses TDM researchers seek to make under this exemption, classic features of fair use. To address opponents’ security concerns, Tait Anderson explained that “reasonable security measures” as used in our petition was concrete enough to require researchers to take precautions to prevent against public dissemination and unauthorized sharing, while not being overly prescriptive in order to accommodate a wide range of TDM projects with different levels of sensitivity in the underlying data. On the topic of existing alternatives to the TDM corpora the TDM researchers seek to compile, Ziyad Alghamdi highlighted the limitations of commercial TDM databases like Hathitrust, which are both limited in the scope of works they contain and how TDM research can be conducted using these works. TDM researchers are seeking this exemption in part because these databases are costly, difficult to use, and incomplete for answering research questions about contemporary literary works and films.

Other topics discussed during the lively hearing included whether the proposed exemption should align with similar carve outs for TDM research in Europe and Japan, how sharing corpora with affiliated researchers for peer review purposes might work, and how and whether literary works and films should be analyzed differently for the purposes of the exemption. The Librarian of Congress is expected to issue a final decision on the proposed exemption in October 2021. We will keep our members and readers apprised of any updates on our proposed exemption as the process moves forward. We’re grateful to law students from the Samuelson Law, Technology & Public Policy Clinic at UC Berkeley Law School for their work supporting our petition for this new exemption.