The One Saving Grace of Google v. Oracle Might be Its Limited Applicability

Photo Credit: iStock/diegograndi

On April 5, the Supreme Court handed down its long-awaited decision in Google v. Oracle, finding that Google’s use of parts of Oracle’s API software code (referred to as “declaring code”) qualifies as fair use. In a 6-2 decision delivered by Justice Breyer, the Court avoided the issue of the software code’s copyrightability entirely—instead assuming that “for argument’s sake” Oracle’s underlying code was copyrightable. As to the second issue at stake in the case—fair use—the Court surprisingly found that all four fair use factors favored Google as a matter of law.

The decision presents a troubling misapplication of the fair use factors and a greater misunderstanding of the goals of the copyright system, but some relief comes in the Court’s explanation that its determination is limited in scope to the specific code at issue in the case and does not “overturn or modify its earlier cases involving fair use.” Despite numerous statements throughout the Court’s decision limiting the applicability of its ruling and analysis to the software code at issue, there is a risk that lower courts will ignore those warnings and wrongly apply this ruling to other types of works, thereby unduly broadening the fair use doctrine more than the Court intended. But we are hopeful that lower courts and commentators will recognize that:

  • The Court’s ruling is an unusual application of the fair use doctrine to a unique form of software code and cannot and should not be applied outside of those specific boundaries.
  • Beginning a fair use examination with factor two distorts the analysis and should not be repeated in future cases, whether they involve computer software code or any other copyright protected work.
  • Merely shifting software code from use in one type of device to another does not constitute a different purpose or character under the first fair use factor, and the opinion wrongly concludes that Google’s use was transformative because the code was used to develop smartphones, rather than desktops or laptops.
  • By focusing on how much of Oracle’s computer code Google didn’t copy, the Court promotes an unsound factor three analysis that is inconsistent with decades of fair use precedent.
  • The Court presents a flawed factor four analysis that disregards the importance of potential markets and would punish copyright owners who don’t quickly and successfully enter into every conceivable market.

After beginning the decision with lines from the Copyright Clause of the U.S. Constitution, Breyer’s opinion takes a strange twist when it says that “[c]opyright encourages the production of works that others might cheaply reproduce.” It’s unclear where Breyer draws this theory from, as no one has ever even remotely suggested that the purpose of copyright is to incentivize the creation of works so that others can cheaply copy them. As if that wasn’t strange enough, Breyer makes no attempt to hide his hostility towards what he considers to be the “negative consequences” of copyright protection likening a copyright owner’s rights to a monopoly and tax. This skewed approach to copyright protection is the prelude to the Court’s fair use analysis, and it’s clear that Breyer’s long history of software copyright skepticism, along with the Court’s general disdain for the Federal Circuit, permeates the entire decision. If not for COVID, this case would have been decided while Ruth Bader Ginsburg was still on the Court. One has to wonder, if she had been, whether Breyer’s software copyright vendetta would have been tolerated.

Factor Two Elevated to Center Stage

In addition to punting on the copyrightability issue, Breyer inexplicably addresses the fair use factors out of order, starting the analysis with the second fair use factor—the nature of the copyrighted work. Many courts have focused on the importance of the first or fourth fair use factors, but this may be the first time any court has ever begun with the second factor first in a way that it controls the entire fair use analysis. Considering the nature of the copied declaring code, the Court found that it is “inextricably bound together with a general system, the division of computing tasks, that no one claims is a proper subject of copyright.”

By beginning the four-factor analysis out of order and finding that the nature of the copied code is similar to other copyrightable materials that serve “functional purposes” and for which protection is “thin,” Breyer digs the Court into a hole out of which a finding of fair use is inevitable. Unlike some recent fair use cases that have seen factors two, three, and four swallowed by transformative determinations under factor one, Breyer’s approach moves the spotlight to the nature of the underlying work, and his assessment that the underlying code is inherently bound to parts of a computer program “that no one claims is a proper subject of copyright” infects the rest of the opinion.

The factor two analysis is also confusing when it says that fair use “can play an important role in determining the lawful scope of a computer program copyright.” In order for a fair use defense to arise, (1) a work must already be determined to be protected by copyright and (2) there must be infringement. While the Court says that it considers Oracle’s code to be protected by copyright for the “sake of argument,” Breyer’s factor two analysis implies that it should not be, and as a result it is never conceded that Google’s use was infringing. It’s a roundabout application of the fair use doctrine that takes on the question of copyrightability, even though the Court claims not to address it.

Continuing with its factor two analysis, the court choses to return to a number of analogies that did not seem to garner any better understanding of the declaring code amongst the Justices during oral argument. Breyer likens the code to “a gas pedal in a car that tells the car to move faster” and also clings to the QWERTY keyboard analogy that he repeatedly raised during oral arguments. But these analogies fail to account for the underlying expressive elements that both sides agree make up the declaring code. The underlying elements of a keyboard—the letters—are not expressive works of authorship, and even if they were, there are a number of different ways a keyboard could be arranged, and copyright protection would not allow the copyright owner to control all alternate keyboard variations.  

The opinion recognizes that Congress specified that computer programs are protectable by copyright, but then goes on to make distinctions based on the perceived “creativity” of the underlying code. The Court finds that the declaring code “embodies a different kind of creativity” than the “creativity” that Google supposedly employed to implement the software in smartphones rather than desktop or laptop computers. It’s a troubling comparison that wades into value judgments of different kinds or levels of creativity—which Congress likely sought to avoid when it amended the Copyright Act to grant protection to computer programs—and it previews the factor one analysis that follows.

Purpose and Character of the Use Found to be Transformative

The opinion then moves into its factor one analysis, admitting that Google precisely copied entire portions of the underlying code “in part for the same reason” they were created. But in considering the transformative purpose and character of Google’s use, the Court finds that the creation of a new platform and new products “was consistent with that creative ‘progress’ that is the basic constitutional objective of copyright itself.” Focusing on the “new computing environment” in which Google implemented the code, the opinion completely fails to address the cogent arguments by Oracle and its amici that the use merely shifts between computing mediums or devices. In fact, it’s astounding that almost nowhere in the opinion does Breyer address the specific arguments made by Oracle and its amici, choosing instead to either take aim at what he sees as errors by the Federal Circuit or to simply explain why he believes Google’s arguments are right. This departure from established judicial practice is striking, and it results in a decidedly one-sided opinion.

As if there were any question given Breyer’s discussion leading up to this, he then finds that Google’s use is a transformative “reimplementation,” which he says furthers the development of computer programs and innovation generally. And yet he makes no mention of the companies that recognized the value of the underlying code and licensed it so that they could build upon it, or the companies that innovated around the Oracle’s code and developed competing platforms and devices of their own. These points were raised repeatedly by different Justices at oral arguments, and by Justices Thomas and Alito in their shrewd dissent, but the only “progress” Breyer’s opinion seems concerned with is Google’s. 

A Flawed Factor Three Analysis

In its factor three analysis, which considers the amount and substantiality of the portion used, the opinion says that the amount of code copied “must be viewed as one small part of the considerably greater whole.” It even goes as far to say that “the better way to look at the numbers is to take into account the several million lines that Google did not copy.” But this approach to factor three invites lower courts to look at how much the infringer didn’t copy –an approach that has been soundly rejected in the past by numerous courts and is wholly inconsistent with decades of fair use precedent.

The opinion goes on to say that Google copied the code because programmers had already learned JAVA’s API system and that it would have been “difficult” to attract programmers to build its system without it. Again, the opinion fails to mention that other companies didn’t need to copy Oracle’s code when they spent an enormous amount of time and resources developing their own code. Instead, Breyer focuses on the popularity and familiarity of the API amongst programmers and the convenience of copying the code, and by doing so effectively penalizes Oracle for the success and popularity of JAVA.

While the majority opinion does not use the word “merger” once, it seems that the Court believes that things like popularity, familiarity, convenience, and attractiveness to users should result in either weakened copyright protection or a more liberal fair use application. These theories have been rightly rejected in the past, and the Supreme Court giving them credence here threatens to create a copyright system in which creating a work that becomes popular and attractive to consumers is punished by weakened copyright protection.

Missing the Mark on Market Effects

The very first line of the fourth factor discussion reproduces language from Section 107 of the Copyright Act, saying that “[t]he fourth statutory factor focuses upon the ‘effect’ of the copying in the ‘market for or value of the copyrighted work.’” But this quote from the statute conspicuously leaves out the word “potential” before “market.” It’s a particularly troubling omission, and likely not a mistake, as the following analysis discounts the potential markets that Sun (or Oracle) could have entered into down the road.

The Court focuses its fourth factor analysis on the public benefit of Google’s use and the “uncertain nature of Sun’s ability to compete in Android’s marketplace.” Apparently, Breyer believes that difficultly entering into a market or an unsuccessful attempt to enter a market means that market no longer exists for the copyright owner and they can’t be harmed by infringers entering into it. He also conveniently glosses over the fact that a licensing market existed and that Google initially sought a license but then abandoned the effort when it didn’t like the terms.

The opinion states that “the record showed that Google’s new smartphone platform is not a market substitute for Java SE.” But that is not what the fourth factor asks. Section 107 clearly states that a fourth factor analysis considers “the effect of the use upon the potential market for or value of the copyrighted work.” The word potential makes all the difference, and yet the majority opinion pays it little attention. To Breyer and the majority, the fact that Sun had difficulty entering into the smartphone market was enough to determine that there was no market effect. It’s a confounding conclusion, and it’s one that would reward an infringer with the resources to race into a market before the actual copyright owner.

While the opinion is limited to a specific type of computer program, it should still be cause for concern among individual creators and small business that own copyrighted works but do not have the resources to enter all possible markets immediately. It is imperative that courts understand that simply because a copyright owner hasn’t successfully exploited a distinct market for a work, it does not mean there isn’t substantial harm done if another comes along, takes your work, and uses it to dominate that market.

Decision is Limited to Specific Forms of Software, But Who Wins?

While this is one of the most perplexing Supreme Court decisions in recent memory—not necessarily because of the result—because of Breyer’s erroneous and poor analysis, the one saving grace is that the decision seems to have very limited applicability. Throughout the opinion, the Court clearly states that the decision “does not overturn or modify its earlier cases involving fair use.” Breyer states numerous times that the decision is limited to the case at hand and only applies to software, and more specifically to declaring code. For example, the opinion says that its conclusion is based on “[a]pplying the principles of the Court’s precedents and Congress’ codification of the fair use doctrine to the distinct copyrighted work here” (emphasis added). The Court repeatedly stresses that the nature of the declaring code is unique, explaining that “this code is different from many other types of code” and “embodies a different kind of creativity.” So despite making distinctions among different types of software—something Thomas’s dissent warns against—and arriving at questionable conclusions in its fair use application, it’s clear that that the Court’s approach is confined to a very specific subset of API code and should not be applied outside of those distinct works.

The limited applicability of the fair use analysis, combined with the Court’s sidestepping around the copyrightability issue, begs the question: What is the long-term value of the decision? Or perhaps more directly, who comes out a winner? Obviously, Oracle and any other companies that own copyrighted software that includes these specific types of declaring code are hurt by the decision. Additionally, the disordered fair use analysis, even if confined to declaring code, is likely to confuse lower courts and non-copyright experts. Also a loser is copyright policy and the public’s understanding of the goals of our copyright system, as Breyer’s focus on only one side of the purpose of copyright misinforms the general public.

The obvious (and only conceivable) winner in this case is Google or similar tech behemoths that have the audacity to build entire platforms on infringement, beat other less powerful and less resourceful companies to a specific market, and spend years (and tremendous amounts of money) defending their actions in court. Some have claimed that this is a win for the software industry as a whole, but it’s not. It’s specific to one company and one company’s particular actions surrounding a very distinct type of copyrighted work. Many who supported Google’s case in amicus briefs and articles wanted the Court to find that the declaring code was not subject to copyright protection, but the Court didn’t do that. Moreover, because the decision reinforces the established principle that fair use is doctrine applied on a case-by-case basis and that its analysis is limited to the specific work and facts at hand, it provides little guidance on fair use generally. The next company that comes along and infringes copyright protected software and claims fair use will probably have to engage in a prolonged legal battle, unless the facts were identical to those in this case. And there aren’t many companies with the resources to do that.

A Sensible Standard of Review

The lone bright spot in the majority decision is the Court’s rejection of Google’s claim that the Federal Circuit’s de novo review of a jury’s fair use finding violates the Seventh Amendment. The question arose after the Federal Circuit reversed a jury’s fair use verdict in favor of Google, and the Court seemed eager to weigh in on the issue, both by requesting supplemental letter briefs by the parties and by devoting significant time to the appropriate standard of review during oral arguments. Google argued that allowing a court to review a jury’s decision de novo, meaning without deference to the jury’s finding, violates its rights under the Seventh Amendment’s Reexamination Clause, which guarantees that “the right of trial by jury . . . be preserved” and forbids courts to reexamine any “fact tried by a jury.”

Rejecting this argument, the Court explains that because fair use is a mixed question of fact and law, and the ultimate question at hand is one of law, “it does not violate the Reexamination Clause for a court to determine the controlling law in resolving a challenge to a jury verdict.” The opinion reminds the parties that this type of review “happens any time a court resolves a motion for judgment as a matter of law,” which is a fact Oracle stressed in its brief and the Assistant Solicitor General raised at oral arguments. The Justices understood that accepting Google’s argument would have upset long-standing motion practice, and the opinion makes clear that “the ‘right of trial by jury’ does not include the right to have a jury resolve a fair use defense.”

Thomas’s Dissent Challenges Majority Opinion

Justice Thomas’s dissent, which was joined by Justice Alito, is highly critical of the majority decision, finding unequivocally that (1) Oracle’s code is copyrightable, and (2) “Google’s use of that copyrighted code was anything but fair.” Thomas explains that declaring code is clearly covered the Copyright Act’s express protection for computer programs as “a set of statements or instructions to be used directly or indirectly in a computer in order to bring about a certain result.” He rebukes the majority’s “wrongful sidestep[ping]” of the copyrightability issue and calls out the opinion for attempting to make a categorical distinction between declaring and implementing code. Importantly, Thomas addresses the merger question, explaining that merger doesn’t apply because there were “innumerable” ways Oracle could have written the declaring code and because other companies such as Apple and Microsoft created their own.

The dissent goes on to question the majority’s distorted fair use analysis, saying that it mistakenly places the nature of the underlying work at the forefront of the analysis in a way that “taints” the entire opinion. Thomas takes particular issue with the majority’s factor four market effects analysis, explaining that Google “ruined Oracle’s potential market” by eliminating the reason manufacturers were originally willing to license or pay for using the code and interfering with Oracle’s ability to license the code for the development of smartphones. He asserts that the majority wrongly focused on the presumed “lock-in” effect Oracle enforcing its copyright might have on the public by possibly limiting future creativity. Once again, Thomas notes that other companies created highly successful mobile operating systems without copying Oracle’s declaring code, and he says that Oracle always made its code freely available to programmers and there is no reason to believe that it would change its practice.

Challenging the majority opinion’s conjecture, Thomas says “[i]f the majority is going to speculate about what Oracle might do, it at least should consider what Google has done.” To Thomas, the Court should be less concerned with a hypothetical situation in which Oracle hinders innovation and more concerned with Google’s recent attempts to monopolize markets and its violations of antitrust laws. Thomas recognizes Google’s unauthorized use as its gateway to market dominance and tens of billions of dollars of revenue, and he warns that [i]f the majority is worried about monopolization, it ought to consider whether Google is the greater threat.”

The dissent also addresses the majority’s flawed transformative analysis, pointing out it “wrongly conflates transformative use with derivative use.” Thomas explains that Google’s use “simply serves the same purpose in a new context,” which is a clear derivative use and not transformative. He reprimands the majority for attempting to create a new definition of transformative, specific to computer programs, that considers whether the use will help others “create new products.” Thomas rightly declares that this type of definition “eviscerates copyright” by implying that someone can get away with pirating the next version of Microsoft Word “simply because he can use it to create new manuscripts.” Quoting from the Court’s Harper decision, Thomas says that rather than transform Oracle’s work, Google “profit[ed] from exploitation of the copyrighted material without paying the customary price.”


In its opinion, the Court reproduces the following line from its 1984 Sony v. Universal decision: “When technological change has rendered its literal terms ambiguous, the Copyright Act must be construed in light of its basic purpose.” To Breyer, the Copyright Act’s purpose lands disproportionally on the side of the secondary user who infringes, so long as there is a perceived public benefit. But access and the public good is only one side of the coin, and the opinion glosses over the critical role copyright protection plays in the commercialization of existing works and the incentivization of the creation of new ones. In doing so, the Court essentially validates Google’s business model of making unauthorized use of other’s intellectual property to build and then solidify its own market dominance. It’s an unfortunate development for a Supreme Court that so rarely takes on fair use, and it’s one that could wreak havoc on the fair use doctrine if applied by lower courts outside of the specific confines of software declaring code. 

get blog updates