Clash of the Technology Titans: Google v Oracle

Gooi Yang Shuh summarises the recent US Supreme Court decision in Google LLC v. Oracle America, Inc. and considers its relevance to Malaysian copyright laws
In what is widely regarded as one of the most significant decisions on copyright in software, the US Supreme Court recently announced its verdict in the case of Google LLC v. Oracle America, Inc. This is the culmination of a long-running dispute between technology giants Google and Oracle, which has been watched closely by many in the tech industry due to the central question which the Supreme Court was tasked to answer: whether or not blocks of codes known as “application programming interfaces” (APIs, for short) are copyrightable, and, if so, the extent to which they can be copied.
To appreciate just why this decision is so significant, we have to firstly understand what APIs are and how they are used.
APIs are lines of code in software programmes that allow multiple software applications to interact and communicate with one another without restriction. In today’s software-driven economy, APIs have become key building blocks to building larger systems. They allow programmers to call upon pre-written computing tasks for use in their own programmes, ensure interoperability between different apps and make it easier to develop complex and integrated platforms.
APIs can be broken down into two categories: “declaring code” and “implementing code”. Implementing codes essentially tell the computer how to execute a particular task; whereas declaring codes call upon the pre-written implementing code to execute that specific task. To illustrate this, Justice Stephen Breyer (who wrote the majority opinion of the Supreme Court) used the analogy of the QWETY keyboard: ‘the declaring code’s shortcut function is similar to the QWERTY keyboard on a typewriter that calls up a certain letter when you press a particular key’.
Thus, the declaring code is the portion of the API which allows easy access to the API’s task library of pre-written codes to select a particular task for implementing complex computing functions using simple commands. Without such access, programmers would have to write their own code from scratch each time to perform the same function.
Returning to the case at hand, Oracle America, Inc. owns the copyright in Java SE, a computing platform that utilises the popular Java computer programming language. Due to a common set of Java APIs, when programmers write software using Java language and the Java SE libraries, the programmer needs to only create one version of his/her software and it can be run on any computing system that supports Java.
In 2005, before the first Android smartphones were launched, Google acquired Android, Inc. and sought to develop the Android operating system for mobile devices. In designing the Android system, Google had independently written its own implementing code. However, to allow programmers familiar with Java programming language to use the API functionality within the Android platform, Google utilised 37 API calls and around 11,500 lines of declaring codes which are part of the Java API.
Hence, in 2010, Oracle sued Google for $9 billion, alleging that Google’s copying of its API codes without licence infringes upon Oracle’s copyright in the same. While Google did not deny copying the Java APIs, it argued that such copying constituted “fair use”, absolving it from any liability for copyright infringement.
After a decade of skirmishes in the trial and appeals Court, the US Court of Appeals for the Federal Circuit held that the ‘structure, sequence and organisation’ of an API can be copyrighted and that Google’s use of Oracle’s API code was not fair use.
Google took this up to the US Supreme Court and presented to the Justices two core questions: (1) whether copyright protection extends to API declaring codes; and (2) if so, whether Google’s copying of portions of the Java API amounted to “fair use”.
Given the potential far-reaching impact which the Supreme Court’s decision could have on the software development landscape as a whole, many industry and legal experts had been following the proceedings with bated breath. A veritable who’s who of the tech industry filed amicus briefs in support of and against both parties’ position. Those in favour of Google included Microsoft, IBM and Mozilla Corporation; whereas Oracle was backed by the Trump administration.
Google and its allies cautioned that an Oracle victory would have a chilling effect on open-source software development, preventing start-ups from using APIs in developing interoperable alternatives through reverse engineering, and forcing companies to implement incompatible standards to avoid risk of infringement. Oracle and its proponents, conversely, warned that a Google victory would weaken copyright protection for software code developers while strengthening the power of Big Tech to take advantage of APIs developed by smaller competitors, reducing the incentive for software innovation. Either way, both sides of the divide argued that a ruling in favour of the other camp would devastate the tech industry.
In April 2021, in a 6-2 majority decision, the Supreme Court ruled in favour of Google and reversed the decisions of the Appeals Court.
The Supreme Court essentially skirted around the first question of whether API codes are copyrightable subject matter and simply assumed ‘purely for argument's sake’ that the entire Java API can be copyrighted. For context, it is well-settled law that software is copyrightable, but Google’s question as to whether discrete portions of software, in this case the declaring code of APIs, are copyrightable is a novel one. Instead, the majority opinion delivered by Justice Breyer focused on the second question of whether Google's copying was fair use and issued a narrow ruling which held that it was.
Generally, “fair use” is an exception and defence to copyright infringement under US copyright law. US copyright legislation statutorily provides four factors for determining whether a particular usage of copyrighted work amounts to fair use: (1) the purpose and character of the use; (2) the nature of the copyrighted work; (3) the amount and substantiality of the copied portion in relation to the copyrighted work as a whole; and (4) the effect the use has on the market value of the copyrighted work.
In his analysis of the four factors, Justice Breyer found that the purpose of Google’s copying was “transformative”, i.e. it ‘adds something new, with a further purpose or different character’ as Google was using the API code to create a different task-related system for smartphones and to build the Android platform. Google’s purpose was, thus, held to be ‘consistent with that creative progress that is the basic constitutional objective of copyright itself’. Further, the amount of code copied (11,500 lines) was found to be one small part of the considerably greater whole – 0.4% of the entire Java API, which consists of 2.86 million lines. As for the fourth factor on the market effect of Google’s copying, not only was it found that the Android OS was not a substitute for the Java SE platform, but it was also found that that Oracle would actually benefit from the reimplementation of the Java SE interface into a different market.
Therefore, all factors taken together, the Supreme Court considered Google’s limited copying of the Java API was fair use as a matter of law. With that, Google dodged a $9 billion bullet.
Reaction among the tech and legal industry has been quite polarising, with some heralding this decision as a victory for the greater software industry and others fearing it further entrenches the monopolistic power that large cash-rich tech companies like Google already have. Then there are those, including this writer, who feel the Supreme Court did not go far enough when it dodged the question whether certain types of code, such as declaring code alone, are copyrightable.
Ultimately, the Supreme Court’s decision is not a radical departure from existing fair use cases and will likely not cause as large of a ripple through the software industry as initially anticipated. While the ruling is arguably narrow and specific to the context of the dispute between Google and Oracle, programmers can breathe a sigh of relief knowing that the commonplace practice of reverse engineering and developing cleanroom versions of existing programmes will not necessarily run afoul of copyright.
Our Malaysian Courts do not typically consider US caselaw as persuasive precedent given that the two countries’ legal systems are fundamentally very different (Malaysian law largely follows that of the UK and other Commonwealth countries). That said, I posit that the decision in Google v Oracle will be highly relevant for future local copyright cases.
Under Section 13(2)(a) the Malaysian Copyright Act 1987, a defence of “fair dealing” for copyright infringement is statutorily provided for. At this point, it must be noted that, traditionally, the difference between “fair use” and “fair dealing” is not one of merely substituting interchangeable synonyms. They are both, in fact, very distinct concepts.
The exception of “fair use” under US copyright law is a very general doctrine that can be given a broad application by the Courts, guided by the four statutory factors in the US Copyright Act, depending on the particular factual matrix of a given case.
“Fair dealing”, meanwhile, is a concept in most common law jurisdictions such as the UK which, unlike “fair use”, is a much more restricted defence which can only be invoked if the use falls within an exhaustive list of statutory fair dealing purposes.
The Malaysian High Court in MediaCorp News Pte Ltd & Ors v MediaBanc (Johor Bharu) Sdn Bhd & Ors [2010] 6 MLJ 657 stated that “Unlike the situation in the United States of America, the [Malaysian Copyright Act 1987] does not allow for 'fair dealing' to be assessed by considering a broad category of circumstances and ascertaining whether those circumstances conform to a set of statutory guidelines. On the contrary, fair dealing under the Act is confined to 'fair dealing' for the prescribed purposes set out in that section and no more.
At that time, Section 13(2)(a) exempted any acts done “by way of fair dealing for the purposes of non-profit research, private study, criticism, review or the reporting of current events” from copyright infringement – akin to the limited “fair dealing” concept in the UK.
However, in 2012, Section 13(2)(a) was amended and now reads “by way of fair dealing including for purposes of research, private study, criticism, review, and the reporting of news or current events”. More significantly, an additional provision was introduced under Section 13(2A) which sets out the exact same four factors as those in the US Copyright Act, for determining whether a dealing constitutes fair dealing.
Unfortunately, there has not been any subsequent caselaw to construe whether Malaysia’s “fair dealing” defence is now actually a general “fair use” defence in disguise. Nonetheless, it is arguable that the amendments to our Copyright Act 1987 have made it such that the High Court’s observations in MediaCorp may no longer hold true, and the Supreme Court’s majority opinion in Google v Oracle (and the entire US body of caselaw surrounding the fair use doctrine) could very well find its way to being cited and relied on in local Courts’ judgments in the near future.
Therefore, the decision in Google v Oracle could provide a helpful framework and guidance to the Malaysian Courts in deciding local copyright infringement cases involving software codes. The “fair use” doctrine may have come to Google’s rescue in this bout, but tech companies should not take this decision to mean that they have carte blanche to make use of others’ API codes without licence.
Case commentary by Gooi Yang Shuh (Associate) of the Intellectual Property Practice of Skrine.