Google v Oracle: A Canadian Viewpoint
April 7, 2021
By Tamara Céline Winegust and Adam Bobker
Earlier this week, the United States Supreme Court (SCOTUS) issued a landmark copyright decision in Google LLC v Oracle America, Inc. Following over a decade of litigation, SCOTUS’s ruling found Google not liable for copyright infringement for copying of 11,500 lines of code from the Java SE computer program, copyright in which is owned by Oracle, into Google’s (then new) Android platform. SCOTUS’s decision turned on whether Google’s actions were “fair use” (they were). The Court specifically did not rule on the copyrightability of Oracle’s asserted code. The decision will no doubt have massive implications for the software industry.
Would the U.S. Court’s reasoning apply directly to similar issues in Canada? Unlikely.
Viewed from a Canadian perspective, SCOTUS’s Google v Oracle decision highlights stark differences between copyright law in the U.S. and Canada, particularly with respect to excusable unauthorized uses (called “fair use” in the United States and “fair dealing” in Canada). Faced with a similar set of facts, the narrower Canadian “fair dealing” claim would likely not apply to excuse the copying. This would mean a Canadian court would be faced with determining the copyrightability of the claimed code. They could not defer the issue to another day, as SCOTUS did.
That said, SCOTUS did raise serious doubts about whether the type of code at issue, namely API declaring code (described further below), was a type of code that was “copyrightable at all.” Although the U.S. Copyright Act protects computer programs, and the Supreme Court specifically did not opine on copyrightability, SCOTUS appears to have been inclined to afford the particular code at issue the thinnest scope of protection, if any, in the context of its fair use analysis. It found Oracle's declaring code “inextricably bound” up with its function and purpose as code—to help divide and organize computing tasks—which themselves are not copyrightable. Such language suggests that if the copyrightability of the API declaring code were considered, the decision may very well have gone in the user’s favour anyway.
The Copied Code
The copied lines of code at issue in Google v Oracle formed part of an “Application Programming Interface” (API), specifically, “declaring code”. APIs are programming tools. They are a package that includes an index of information that sets out, in an organized manner, a variety of available computing tasks and how to invoke them. The “declaring code” is essentially a named template that programmers can use to invoke a computing function, the implementation of which function involves different code. Java’s programming language is widely used by programmers. The evidence showed Google copied the Java API declaring code for 37 packages to facilitate uptake by programmers who were already familiar with the Java programming language, which would then ensure widespread development of programs for Android systems. However, Google developed its own code to implement the tasks themselves.
The SCOTUS decision …
Google’s unauthorized copying of code was ultimately excused by SCOTUS as a “fair use”. This defense to copyright infringement appears in the U.S. Copyright Act itself, in section 107. That section sets out that “fair use of a copyrighted work … for purposes such as criticism, comment … is not an infringement of copyright”, as well as that factors are to be considered when evaluating whether the use in question is “fair”, including four specific factors that “shall” be considered ((a) the purpose and character of the use; (b) the nature of the copyrighted work; (c) the amount/substantiality of the amount taken; and (d) the commercial/market effect of the use).
In Google v Oracle, Justice Breyer, writing for the majority of judges, confirmed a long line of decisions that interpreted statutory fair use in section 107 as being essentially open—the types of permitted use are not a closed universe (due to the wording “such as” in the preamble), and the factors are not exhaustive (although the four enumerated factors must be considered). For the majority of SCOTUS, important considerations in approaching the fair use evaluation included (a) the purpose of copyright protection as written in the U.S. Constitution—to “promote the Progress of Science and useful Arts”; and (b) the doctrine’s “judge-made origins”, which demonstrate the flexible nature of fair use to address contexts where copyright protection is “thin”.
Although proceeding to consider each of the four fair use factors, the driving principle behind the Court’s analysis appears to be whether Google’s use was “transformative”—i.e., whether its use was something that “that adds something new and important”; for example, a new expression, meaning, or message. In this case, such “transformation” was found by Google’s use of the code to create “new products”, including “a highly creative and innovative tool” for programmers in the smartphone environment. This use was “consistent with the creative ‘progress’ that is the basic constitutional objective of copyright itself”, and thus weighed heavily in favour of fair use. The commercial nature of Google’s use did not change this calculus. Moreover, because the amount of code copied by Google was “tethered” to the valid (and transformative) purpose, the “amount/substantiality” factor weighed in favour of fair use as well. Last, with respect to the “market effects”, the Court noted that given the widespread adoption of the Java programming language by programmers, over broad protection of Oracle’s declaring code could limit “the future creativity of new programs.” This would “interfere with, not further, copyrights basic creativity objectives”, even if such control were financially lucrative.
… and Canadian copyright law
The concept of “Fair Use” does not exist under the Canadian Copyright Act. Instead, the Canadian Copyright Act provides for a concept called “fair dealing”. While it shares some similarities with “fair use”—particularly, enumerated factors to consider when evaluating the fairness of the dealing—there are important differences. Unlike the U.S. doctrine, which has an “open” concept of permitted uses, the acceptable dealings under the Canadian Copyright Act are specific and closed. To qualify, a dealing must be for the purposes of research; private study; education; parody, satire; criticism or review; or news reporting. In fact, until amendments to the Canadian Copyright Act in 2012, “education”, “parody”, and “satire” were not enumerated dealings. Although over the last two decades, the Supreme Court has emphasized that the enumerated categories should be granted a “large and liberal interpretation” (with the “conceptual heavy hitting” occurring when “fairness” is evaluated), courts have been clear that unless a dealing falls within one of the enumerated categories, a “fair dealing” claim will fail.
Moreover, unlike U.S. copyright law, which incorporates some doctrines of “judge-made origin” (namely fair use), copyright in Canada is entirely a creature of statute. The rights and remedies afforded under it are exhaustive. As shown in the context of fair dealing, it can (in some contexts) lack the flexibility that can facilitate organic judicial, rather than structured legislative, intervention to address technological and social developments around uses.
In addition, unlike in the United States, copyright has no higher constitutional purpose in Canada. The Canadian constitution simply (and unceremoniously) designates “copyright” as subject 23 in a list of 29 subjects over which the Federal government has power (The Constitution Act, 1867, 30 & 31 Vict c 3, at s 91). It has been left to the courts to develop the principles underlying the purpose of copyright in Canada. This has arguably led to a more balanced user-owner approach compared to the United States—the Supreme Court has repeatedly held that the overarching purpose of the Canadian Copyright Act is to both promote the public interest in the encouragement and dissemination of artistic and intellectual works, and justly reward the creator of the work, and has been developing principles of “technological neutrality” in recent decisions.
Considering Google v Oracle in the Canadian context, there are doubts whether Google's copying of API declaring code could be excused as a “fair dealing”—not because such dealing could not be considered fair, but rather because such dealing may not fall under one of the enumerated purposes (even with a “large and liberal interpretation”).
There is also a question of whether Google’s dealing with the API declaring code would be considered “fair” in Canada, even if it could fit into an allowable purpose. At this time, Canadian courts have not recognized “transformative use” as a characteristic of fair dealing. One British Columbia court noted that “what may be transformative, and as a result fair use in the U.S., may still be copyright infringement in Canada." (see Century 21 Canada Limited Partnership v. Rogers Communications Inc., 2011 BCSC 1196 at para 234). In view of the centrality of the transformative nature of Google’s use of the Java API declaring code, it is certainly open to debate whether a “fairness” analysis that omits this consideration could result in a similar conclusion that the dealing was “fair”.
Last, while the Canadian Copyright Act does contain specific exceptions for copying computer programs, namely for compatibility, backup, interoperability, encryption research, and security, such exceptions include very specific qualifying language and are unlikely to apply to the copying of a part of code that is incorporated into another program. For example, the language of the interoperability exception requires the copy is made for the “sole purpose of obtaining information” that would facilitate interoperability.
Although both Canadian and American copyright statutes provide specific protection for computer programs, the particular facts and reasons underlying the United States Supreme Court’s Google v Oracle decision raises questions about whether a similar decision would issue in Canada. In particular, differences as between the fair use doctrine in the United States and fair dealing claims in Canada could well operate to produce a different outcome—one that may not have excused Google’s activities if carried out here. Instead, in Canada, the case may need to rest on whether the unauthorized copying was of a “computer program” that benefits from copyright protection. SCOTUS provided some indication as to whether it believed Oracle’s copied API declaring code was protectable under copyright—the Court expressed doubt, particularly since the API declaring code was “inextricably bound” up with its function and purpose as code. However, the question of copyrightability of API declaring code was not definitely answered as a matter of law by the U.S. Supreme Court. If given the opportunity, Canada’s courts may well be forced to do so.
Content shared on Bereskin & Parr’s website is for information purposes only. It should not be taken as legal or professional advice. To obtain such advice, please contact a Bereskin & Parr LLP professional. We will be pleased to help you.