L’Association des archivistes du Québec

Les informations de ce site Web à été fournie par des sources externes. Le gouvernement du Canada n'assume aucune responsabilité concernant la précision, l'actualité ou la fiabilité des informations fournies par les sources externes. Les utilisateurs qui désirent employer cette information devraient consulter directement la source des informations. Le contenu fournit par les sources externes n'est pas assujetti aux exigences sur les langues officielles et la protection des renseignements personnels.

September 16, 2021

Archival Community Response to “A Consultation on a Modern Copyright Framework for Artificial Intelligence and the Internet of Things”

Presented by the Canadian Council of Archives Endorsed by the Association of Canadian Archivists and l’Association des archivistes du Québec

Canadian Council of Archives (CCA) Statutory Review Working Group
Email: Nancy.Marrelli@concordia.ca

Table of Contents

                                                                                                               

Executive summary

We have carefully considered and commented on the relevant issues presented in the consultation paper on Artificial Intelligence and the Internet of Things in Canada.  The important public service mission of Archives as a trusted repository for purposes of research, education, and culture, must be taken into careful consideration in implementation of any changes to the copyright framework. 

Recommendations

Our recommendations include:

  • Ensure public policy and legislation that involves AI is undertaken only after wide public consultation and discussion that includes broad public policy concerns such as protection of privacy and personal information, and other human rights.
  • Add an exception to permit the making of copies of legally acquired material in order to carry out a computational analysis of the content of the material for commercial and non-commercial purposes. The exception must not be nullified by contractual terms.
  • Outputs of AI process should be in the public domain and should not receive full copyright protection. The terms of any protection that may be assigned to AI outputs should receive full public discussion and debate, creating no more than a variation of what is sometimes called “thin copyright”.
  • Circumvention of TPMs should be permitted for any activity that is otherwise allowable under the Copyright Act.
  • Amend the Copyright Act to allow LAMs and all Canadians to circumvent TPMs in order to diagnose, maintain, or repair their products, not subject to contract override.

Introduction

This submission is made on behalf of the Canadian Council of Archives in response to the Government of Canada’s consultation paper on a Modern Copyright Framework for Artificial Intelligence and the Internet of Things. The submission has been fully endorsed by l’Association des archivistes du Québec, and the Association of Canadian Archivists.

Copyright is of great importance to the archival community. We are pleased to participate in this consultation and we look forward to a further opportunity to comment on and respond to all related legislative initiatives.

Background

Archives are non-profit institutions with an important public service mission: to acquire, preserve, and provide access to our documentary and cultural heritage as a trusted repository for researchers and users in Canada and world-wide for purposes of research, education, and culture.

The entire Library, Archives, and Museum (LAM) community shares a commitment to its public service mission for the sharing of library, archives, and museum holdings. LAMs make our holdings available for public access, as a public service, and not for commercial gain.  LAMs have a long and honoured tradition of providing access to Canadian heritage, and archives provide a safe haven for access to the archival holdings that are the source of information, inspiration and innovation for many creators as they develop new works in a variety of formats. Like libraries and museums, archives are committed to promote the works of Canadian creators through many public-facing programs.

It is clear that Artificial Intelligence and the Internet of Things is introducing significant disruption to current realities, and it is important to remember that this disruption will take place not only in the commercial marketplace, but also in research, and far-reaching public policy issues that affect the daily life of ordinary citizens.

What is Special About Archives?

Archives are the materials created or received by a person, family or organization in the conduct of their affairs. They are preserved because of the enduring value of the information they contain or as evidence of the functions and responsibilities of their creator. The Universal Declaration on Archives clearly expresses the public interest role that Archives fulfill:

“Archives are a unique and irreplaceable heritage passed from one generation to another. ... They are authoritative sources of information underpinning accountable and transparent administrative actions. They play an essential role in the development of societies by safeguarding and contributing to individual and community memory. Open access to archives enriches our knowledge of human society, promotes democracy, protects citizens’ rights and enhances the quality of life.”Footnote 1

It is useful to note that archives and archivists serve both the creator and the user stakeholder communities. Many of our donors are creators and we have both legal and ethical obligations to safeguard their materials and to respect their intellectual property rights in all of our activities. We also serve the user community by acquiring, preserving, and making accessible, the documentary heritage of Canada.

The nature of archival material presents particular challenges from the perspective of copyright. The combination of vast numbers of works and diverse other materials, most of them unpublished, the fact that most were not created for commercial purposes, the high proportion of orphan works, and the absence of licensing models means that solutions to many difficult issues are not obvious. Archivists work with paper, film, audio, and digital records. Most archival materials are the accumulated records of governments, companies, charities, families and individuals. They consist of works like letters, diaries, emails, financial statements, minutes of meetings, reports, photographs, audio recordings, home movies, and increasing numbers of data sets. Each letter, photo or report is distinct and the extent is vast – it is not an exaggeration to say that Canada’s archives hold billions of works.

Archival holdings serve as the raw material for countless new works in all formats. In order to achieve a satisfactory public policy resolution, any attempt to develop a modern copyright framework must include consideration of the particular characteristics of archival material and how such a framework would impact the public mission of archives.

In this consultation response, our remarks are largely limited to matters relating to Archives and the LAM community. We leave it to other stakeholders to respond in detail to issues related to their mandate.

Further Public Consultation Must Extend Beyond Copyright Stakeholders

LAMs provide equitable access to their holdings, without discrimination and as an essential public service that is independent of the marketplace and vested interests.  Archives are trusted repositories that facilitate global access to Canadian documentary heritage and the vital information that it contains, serving the public interest.

Artificial Intelligence is introducing disruption to current realities in the marketplace, in research, and in many other spheres of public life. Society is struggling to understand the impacts of these rapidly evolving technologies. Despite the many benefits of artificial intelligence, as the consultation paper noted, it has also raised several public interest issues that are of concern to LAMs and that go well beyond copyright concerns, such as machine learning bias, informed consent for the use of data, privacy issues, data breaches, and protection of freedom of expression. The Archival Community believes that all these issues require careful examination and public consultation that extends beyond this consultation.

Any changes to policy and legislation, be it copyright or other initiatives, must safeguard the public interest and privacy considerations. Effective public consultation with public interest groups and the wide general public on the complex issues surrounding AI must take place. It is not only copyright stakeholders and those with a vested interest in the marketplace that should be part of the discussion. Copyright concerns for AI must not be addressed in isolation and without reference to broad public concerns such as protection of privacy and personal information and other human rights. These are extremely important issues for archivists because of the nature of our holdings and our public service mandate. Broad public consultation and discussions are required to develop sound public policy and legislation that will avoid decision-making that is solely market-oriented. The public good will be served by such consultations that are in line with Open Government policies and processes already in place in Canadian government.

Recommendation:

Ensure public policy and legislation that involves AI is undertaken only after wide public consultation and discussion that includes broad public policy concerns such as protection of privacy and personal information, and other human rights.

Some Issues Presented in the Consultation Paper

Text and Data Mining

Text and data mining (TDM) comprises various technologies for the computer-based analysis of large amounts of copyright works to discover patterns, trends and other useful information that cannot be detected through human reading. TDM holds great potential for research and innovation. However, to ‘mine’ texts, images, and other content, researchers need to access, copy, and process them using computer programs. Even if researchers can lawfully access and read the material, copying a substantial part of works may infringe copyright in those works.  But copyright was never meant to restrict the use of the ideas, facts and information that exist in a work. In the analogue world, merely viewing or reading it is not an infringement. TDM simply employs computers to “read” material and extract facts – something that humans already have the right to do. Thus, copying for text mining is a “non-consumptive use” that is not intended to exploit the copyright protected material, and should be considered merely incidental to how TDM works.

At present, archival institutions hold relatively few research data sets that will be the raw material for TDM. Increasingly, however, archives will continue to acquire such materials from their parent institutions as well as private donors. As well, archives already contain traditional records that, when digitized, can be mined for historical information. For example, fur traders’ journals document decades of weather patterns, and county atlases record names of residents.   Fundamental to the archival mission is making our holdings available for research by our users, typically done by making a copy for the user. Archival institutions already are the beneficiaries of several exceptions that permit making copies for users, namely fair dealing, and s. 30.21, which permits making a single copy of unpublished works for the purposes of research and private study, subject to certain conditions. However, it is not entirely clear how these provisions apply to TDM, thus hindering our mission and the research that is dependent upon it.

Thus, an exception that extends beyond LAMs is needed to support TDM. Such an exception would permit researchers to make copies of legally acquired material for the purpose of carrying out a computational analysis of anything recorded in the material. The exception would apply to all categories of works and other subject matter, and would permit research for both commercial and non-commercial purposes. Contractual terms that restrict or prevent the acts permitted under the exception would be unenforceable.

Recommendation:

Add an exception to permit the making of copies of legally acquired material in order to carry out a computational analysis of the content of the material for commercial and non-commercial purposes. The exception must not be nullified by contractual terms.

Authorship

In Canada copyright is automatic for original creations that include skill and judgement. Can an AI generated creation meet this originality threshold? The humanly created algorithm that is used to generate the AI creation does meet the threshold and can definitely be protected by copyright, but the creation that results from the AI mechanical process cannot meet this requirement for skill and judgement.  Assigning full intellectual property rights to the output of AI processes is inappropriate. We might consider certain very limited rights for AI outputs, a variation of what is sometimes referred to as “thin copyright”, such as the limited rights sometimes accorded to databases, but these should be very limited in both scope and duration, and subject to careful deliberation.

Whatever regime is adopted for AI generated outputs will affect creators. Assigning full copyright protection to AI output would very negatively affect how society and creators view the work of creators, and their contribution to society, resulting in a negative effect on incentive to create. It is crucial to not equate human and machine-generated creation. The significance of this consideration should not be underestimated.

In addition, we must be wary of assigning full intellectual property rights to outputs of AI processes that can be programmed to create mass output that could quickly monopolize the creative space and disrupt both copyright balance and the marketplace.

Recommendation:

Outputs of AI process should be in the public domain and should not receive full copyright protection. The terms of any protection that may be assigned to AI outputs should receive full public discussion and debate, creating no more than a variation of what is sometimes called “thin copyright”.

Technical Protection Measures and Right to Repair

The Copyright Modernization Act of 2012 introduced Technical Protection Measures (TPMs) in a way that was too narrow from the outset and has become increasingly outdated in the past 10 years. The legal prohibition against the circumvention of TPMs was introduced as a form of incentive for rightsholders to make their works available in digital format; this reasoning shows just how dated the measure is. In 2021, distribution in digital formats is ubiquitous across creative outputs. Perhaps in 2012 this legislation applied to outlier content that was at a higher risk of infringement; however, the illegality of circumventing TPMs is now causing widespread and undue restrictions on the primary formats in which copyrighted works are consumed.

TPMs become a major problem when they undermine the public interest mandate of LAMs by limiting otherwise non-infringing activities. Circumventing a TPM is prohibited even to make a copy for preservation or collection management purposes. It means that archivists cannot carry out the essential activities required to ensure the long-term preservation of, and access to, important parts of our documentary heritage that are constrained by TPMs. Non-infringing activities, such as reformatting documents, photographs or maps that are in an obsolete format, or a format that is in danger of becoming obsolete, have long been recognized in the Copyright Act as being both necessary and allowable for LAMs. Although CUSMA limits TPM circumvention to specific purposes, it must be recognized that these exceptions do not address a variety of legitimate activities carried out by archives.

Not only has the TPM provision upset the copyright balance in favour of rightsowners, but it has also extended copyright protection to a whole new domain through restrictions on the “right to repair”. Previously, consumers might run the risk of a void warranty for attempting to fix their own devices; with the introduction of TPMs in the Copyright Act, such a routine activity becomes copyright violation. Now, when a device containing a TPM breaks or malfunctions, any attempt by the consumer to repair the device may result in a copyright violation if the repair requires reproducing a substantial part of a copyrighted work (e.g., software) or circumventing a TPM. This poses a major obstacle to archives for which in-house maintenance and repair is often the only option available to ensure ongoing access to holdings. In addition, archives tend to house unique and rare materials for which copyright owners are unlocatable to assist in repairs; resulting in archival holdings which are inaccessible and potentially irreplaceable.

Recommendations

Circumvention of TPMs should be permitted for any activity that is otherwise allowable under the Copyright Act.
Amend the Copyright Act to allow LAMs and all Canadians to circumvent TPMs in order to diagnose, maintain, or repair their products, not subject to contract override.

Conclusion

The Archives community has examined the implications of potential changes to the Copyright Framework for Artificial Intelligence and the Internet of Things from the archival perspective. We believe there are many public benefits to AI and it is important to take appropriate actions to ensure Canadians can move forward with changes relating to TDM and TPMs/Right to Repair in order to be able to compete in research and in the marketplace in the current environment where AI is a reality.  But it is also important to carefully examine and take into consideration all the other related public policy issues, some of which are not directly related to copyright before developing public policy and legislation around the complexities of AI.

The Archival Community believes that human creation and AI outputs do not merit the same copyright protections. Any protections assigned to AI outputs must receive careful consideration, and where merited, they must be very limited. We believe it is important to add an exception to the Copyright Act to permit the making of copies of legally acquired material in order to carry out a computational analysis of the content of the material for commercial and non-commercial purposes. As we have stated on numerous occasions, we believe it is crucial that circumvention of TPMs should be permitted for any activity that is otherwise allowable under the Copyright Act.  In addition, all Canadians should be able to circumvent TPMs in order to ensure the Right to Repair.