Professor Michael Geist

The information on this Web page has been provided by external sources. The Government of Canada is not responsible for the accuracy, reliability or currency of the information supplied by external sources. Users wishing to rely upon this information should consult directly with the source of the information. Content provided by external sources is not subject to official languages and privacy requirements.

Government of Canada Consultation on a Modern Copyright Framework for Artificial Intelligence and the Internet of Things

Submission by Professor Michael Geist, Canada Research Chair in Internet and E-commerce Law, University of Ottawa, Faculty of Law
Centre for Law, Technology and Society

September 2021

A. Overview

I am a law professor at the University of Ottawa where I hold the Canada Research Chair in Internet and E-commerce Law and serve as a member of the Centre for Law, Technology and Society. I focus on the intersection between law and technology with an emphasis on digital policies. I have edited multiple texts on Canadian copyright law and appeared many times before House of Commons committees on copyright law and policy. I submit these comments in a personal capacity representing only my own views.

My submission focuses on the benefits of flexibility within copyright law and the need to ensure that Canada’s copyright legal framework strikes an appropriate balance between creator rights and user rights as repeatedly emphasized by the Supreme Court of Canada. I believe that the 2012 copyright reforms made the critical mistake of restricting innovation by not including a flexible fair dealing provision and by implementing restrictive anti-circumvention rules that extended far beyond international requirements.

The current consultation represents an opportunity to remedy those errors by adopting a flexible fair dealing provision (the “such as” approach) and linking the anti-circumvention rules more directly to actual copyright infringement. Should the more flexible approach not be adopted, I would recommend a new fair dealing exception for text and data mining to support artificial intelligence and new anti-circumvention exceptions that explicitly establish rights related to interoperability and repair.

I further note that these recommendations are broadly consistent with the recommendations of the 2019 copyright review. While the consultation document also references the Canadian Heritage supplemental study, it notes that “it heard less stakeholder testimony about these challenges and did not make any recommendations regarding AI and IoT.” I have written extensively about how that committee’s work did not consult widely and that the INDU study should be viewed as authoritative.

B. Fair Dealing: Flexible Fair Dealing and a Text and Data Mining Exception

The consultation raises the possibility of an exception for text and data mining. I believe that the better approach would be greater fair dealing flexibility – adopting the “such as” approach – which would make the current list of fair dealing purposes illustrative rather than exhaustive and would place Canadian innovators on a level playing field with fair use countries such as the United States.

Led by the U.S., several countries around the world, including Israel, South Korea, and Singapore, have established fair use provisions within their copyright laws. Fair use does not mean free use – rather, it means that there is a balance that allows certain uses of works without permission so long as the use is fair. A “such as” fair dealing reform would still maintain a full fairness analysis along with longstanding jurisprudence to minimize uncertainty.

This recommendation is consistent with the unanimous opinion of the 2019 copyright review, which concluded:

Parliament should make the list of purposes enumerated under section 29 of the Act an illustrative list rather than an exhaustive one. Doing so would increase the flexibility of the Act by allowing a broader range of admissible purposes to emerge from existing ones under the guidance and the supervision of the courts—for example, from criticism to quotation, from parody to pastiche, and from research to informational analysis. Such an amendment could allow new practices to fall under fair dealing, such as “reaction videos” and video game streaming. The Committee emphasizes that the purpose of a dealing is only one of many factors taken into account when determining whether this dealing is indeed fair under section 29 of the Act.Footnote 1

Consistent with that recommendation, the best approach for addressing the innovative potential of artificial intelligence and the concerns that restrictive fair dealing could limit Canada’s competitiveness in the field, is to establish a flexible fair dealing model that can be readily adapted to technological change.

In the absence of a flexible fair dealing approach (or in addition to such a change to allow for greater certainty), an explicit exception for text and data mining (sometimes referred to as informational analysis) is needed within the Copyright Act’s fair dealing provisions. The inclusion of a specific text and data mining exception is also consistent with the 2019 copyright review recommendations:

The evidence persuaded the Committee that facilitating the informational analysis of lawfully acquired copyrighted content could help Canada’s promising future in artificial intelligence become reality. The Committee therefore recommends:

Recommendation 23

That the Government of Canada introduce legislation to amend the Copyright Act to facilitate the use of a work or other subject-matter for the purpose of informational analysis.Footnote 2

The federal government has invested millions to support research and commercialization of AI in the hopes of making it a world leader. They hope that by attracting high-profile talent and significant corporate support they can transform strong AI research into an economic powerhouse. However, the current state of Canadian copyright law undermines this investment by inhibiting innovation through the creation of legal uncertainty and high barriers faced by the very groups the investment aimed to attract.

AI research works by making machines smart. Whether this is focused on automated translation, big data analytics, or new search capabilities, it is dependent on data being fed into the system. Machines learn by scanning, reading, listening, or viewing human created works. The better the inputs, the better the outputs and the more inputs there are, the likelihood that results are biased or inaccurate decreases.

For example, in 2016 the United Nations released 800,000 manually translated documents in the six official UN languages for machine use. These perfectly translated data sources in multiple languages helped create better automated translation systems. The one downside is treaties and diplomatic correspondence rarely mimics everyday speech. These systems would benefit from a range of translated materials – such as popular books or television shows. This shows there is significant value in a breadth of possible inputs to be used when conducting AI research. Restrictive copyright law, such as in Canada, hinders this.

Canadian copyright law inhibits this because restrictive rules limit the data sets that can be used for machine learning purposes, resulting in fewer pictures to scan, videos to watch, or text to analyze. Without a clear rule to permit machine learning, the Canadian legal framework trails behind other countries that have reduced risks associated with using data sets in AI activities. Under the Canadian system, researchers either must risk copyright infringement by using protected works to make their machines smarter (which has a chilling effect on innovation), or severely limit the data sets used, thereby producing less “smart” machines than would be possible under a more open copyright regime. This raises concerns of bias and discrimination.

Within the current framework, the fair dealing rules provide some protections and allow for some use of copyrighted work by AI companies without permission. Canadian courts have ruled that it is a right that should be interpreted in a broad and liberal manner and there are several purposes that would permit some text and data mining activities – notably exceptions for research, education, and private study. However, the Canadian government’s emphasis on the commercial benefits of AI make it so that fair dealing likely will not offer sufficient flexibility to move from research done in a lab or classroom to a commercial product. There is no clear fair dealing exception for the commercialization of outputs that used copyrighted material. Therefore, there needs to be a broad exception for text and data mining in the Copyright Act that encompasses both research and commercial works.

The corporations and high-profile talent attracted by the investment in the Canadian AI system have been calling for such an exception as well. In 2018, various technology groups, such as Element AI based out of Montreal, note that the current uncertainties in the Copyright Act limit the ability for Canadian companies to “access a basic necessary resource to train their algorithms”.Footnote 3

Internationally, other countries have addressed this exact issue through text and data mining exceptions. In the above-mentioned discussion, Microsoft noted that there was a broad acceptance of text and data mining exceptions around the world and that Canada is posed to fall behind and be at a global disadvantage without one. They also noted that, unsurprisingly, the countries that have/are considering these exceptions are at the forefront of research involving data analytics and artificial intelligence.

Examples of such exceptions can be found in Japan, the European Union, and the United Kingdom. The EU enacted a mandatory exception for text and data mining for the purposes of scientific research but has permitted rightsholders to “contract out”.Footnote 4 The U.K.’s exception allows copies of works to be made without permission of the copyright owner for the purposes of automated analytical techniques to analyze text and data for patterns, trends, and other information. The law does not allow contracts to restrict data mining activities, but the exception is limited to non-commercial research.

In order to not fall behind internationally and position ourselves as a world-leader, Canada needs to adopt a broad exception for text and data mining, including both research and commercial uses. Ultimately, machine learning does not harm the primary purposes of the original work – the goal is not to republish or compete with copyrighted materials, but to ensure that researchers and AI companies can mine the text and data for informational analysis purposes – thus including commercial uses in the exception will not harm rightsholders and will facilitate Canadian innovation. The fact that other countries have limited the exception to research purposes gives Canada an opportunity to become an economic leader in AI, but only if they adopt a broad exception.

C. Technological Protection Measures

Canadian anti-circumvention laws are among the strictest in the world and stifle innovation. Technological protection measures (also known as digital locks) are designed to make it difficult for users to copy works in digital formats. Copyright law grants legal protection to these measures, through anti-circumvention rules, by making it an infringement to simply “pick” or break the digital lock (in some cases, it is even an infringement to simply make tools or devices that can be used to break the lock). Digital locks cause a problem, however, when these legal provisions are used in cases that have nothing to do with copyright.

Anti-circumvention legislation is required under the WIPO’s (World Intellectual Property Organization) Internet Treaties, but it leaves considerable flexibility in how they should be implemented.Footnote 5 This is reflected in the countless examples around the world of countries adopting flexible anti-circumvention rules. Countries like New Zealand, Switzerland, and Germany have tied their anti-circumvention legislation to copyright to ensure that non-copyright infringing circumvention is still allowed. For example, the New Zealand law does not apply to devices that “only control access to a work for non-infringing purposes” and only punishes selling circumvention devices if the person knows or has reason to believe it will be used to infringe copyright. It clearly permits circumvention for “permitted acts” such as fair dealing rights and also created a system to facilitate circumvention for permitted acts by a qualified person if the user is unable to circumvent the TPM themselves.

The United States lobbied for more stringent anti-circumvention requirements during the drafting of the WIPO Internet Treaties, but this was rejected for the current model. However, the domestic U.S. system contains rather strict rules, and in 2012 it successfully pressured the Canadian government into adopting some of the strictest anti-circumvention measures in the world, stricter than what the U.S. has domestically. For example, the U.S. has a system that reviews every three years whether new exceptions are needed. This has led to the adoption of several new exceptions to TPMs for innovative activities such as automotive security research, repairs and maintenance, archiving and preserving video games, and for remixing from DVDs and Blu-Ray sources.Footnote 6

From a Canadian perspective, the best policy approach would be to clarify that the anti- circumvention rules do not apply to non-infringing uses. This would enable the anti- circumvention rules to work alongside the user rights in the Copyright Act (also known as limitations and exceptions) without restricting their lawful exercise. This approach was endorsed by the 2019 copyright review report:

However, it agrees that the circumvention of TPMs should be allowed for non-infringing purposes, especially given the fact that the Nintendo case provided such a broad interpretation of TPMs. In other words, while anti-circumvention rules should support the use of TPMs to enable the remuneration of rights-holders and prevent copyright infringement, they should generally not prevent someone from committing an act otherwise authorized under the Act.Footnote 7

Adopting this approach would effectively address ongoing concerns about restrictive anti- circumvention rules and reduce the need for regular reviews of the law to identify instances where it is unduly restrictive for innovation as well as harmful to the rights of users. In the absence of this preferred approach, specific exceptions for interoperability and right of repair is needed.

  1. Interoperability

Interoperability exceptions to the anti-circumvention measures can be found in the Copyright Act, but they do not go far enough to maintain a level playing field. A classic example involves the use of Linux as a consumer operating system. This operating system cannot officially play DVDs since most commercial DVDs contain a digital lock and the entity that controls the lock does not license the necessary locks to play DVDs on Linux. Programmers have developed alternatives, but all involve circumventing the digital lock. The interoperability provisions do not help address this issue, since DVDs may not be considered computer programs and many of the circumventing programs have functionality beyond playback of commercial DVDs. The net effect is that anti-circumvention measures erect an enormous barrier to open-source software adoption, thereby harming innovation and a competitive marketplace. The solution is to create an exception the substantially broadens the interoperability exception.

The Nintendo case also shows the limits of the Canadian interoperability exceptions. The court rejected the modchip seller’s arguments that the modchips that were being sold allowed users to play “homebrew” games that are designed to be used on Nintendo consoles but are not licensed by the company. The interoperability exceptions could feasibly be applied to such games, but this was rejected by the court, saying that homebrew usage was overshadowed by infringing activity.

Thus, the lack of expansive interoperability exceptions harms Canadian innovation because they place barriers to the ability of third parties to develop new software that is interoperable with existing programs.

  1. Right of Repair

The lack of a right of repair exception in Canadian digital lock rules also hinders Canadian innovation significantly, as best exemplified by the current problems faced in the agriculture and protein industries. These industries require increasingly sophisticated farm equipment with many of the same features as video game consoles or other software-enabled devices. Copyright rules have been leveraged by manufacturers to limit farmer’s “right of repair”.

The anti-circumvention measures may prohibit Canadian farmers from tinkering with their equipment to create new innovations or simply to repair malfunctioning equipment. As Carlo Dade of the Canada West Foundation noted in a December 2018 op-ed, “Currently, in the middle of critical harvest time, a farmer in Emerson, Man., can look across the border to Pembina, N.D., and see a farmer there hack her tractor to fix a problem without consequence. But should the Manitoba farmer try this she would face serious sanction. Or more starkly, the Manitoba farmer could drive their tractor across the border to hack the software without problem or sanction, but should she drive the tractor back into Canada she would be driving into trouble.”Footnote 8

After farmers protested against similar copyright restrictions, the U.S. established specific exceptions permitting digital locks to be circumvented to allow repair of software-enabled devices. The new U.S. exception is part of a gradual loosening of the digital lock rules in that country including new exceptions for automotive security research, repairs and maintenance, archiving and preserving video games, and remixing videos from DVDs and Blu-Ray sources.

Given the impact on consumers, the agricultural sector, and the environment, a provision that explicitly permits circumvention for purposes of the right of repair in Canada is long overdue. Indeed, such an approach is consistent with the 2019 copyright review recommendation:

Recommendation 19

That the Government of Canada examine measures to modernize copyright policy with digital technologies affecting Canadians and Canadian institutions, including the relevance of technological protection measures within copyright law, notably to facilitate the maintenance, repair or adaptation of a lawfully-acquired device for non-infringing purposes.Footnote 9