Google v. Oracle at the Supreme Court: Copyrightability, Fair Use, and Standard of Review

The following post comes from Chris Wolfsen, a recent graduate of Scalia Law and a Research Assistant at CPIP.

By Chris Wolfsen

Grocery store shelves, QWERTY keyboards, and restaurant menus. These are just three of the analogies that Supreme Court justices used to grapple with the complex issues in the long-awaited Google v. Oracle oral argument that was heard last Wednesday, October 7. The case between the two tech companies that has been brewing for over a decade comes down to 11,000 lines of Oracle’s Java code that Google copied, without permission, for use in its well-known Android smartphone platform—a competing product.

Oracle originally filed suit in 2010 against Google in the U.S. District Court for the Northern District of California for both patent and copyright infringement. In 2014, the Federal Circuit ruled that Oracle’s code was copyrightable, and Google appealed to the Supreme Court, only to be denied review. On remand, again in the district court, Google claimed that its use of the Oracle code was fair use—meaning that even if it were an infringing use it should still be allowed by law. In 2016, a jury sided with Google, and Oracle filed a motion to challenge the jury verdict, which was denied. So instead, Oracle appealed the case back up to the Federal Circuit where the court reversed the lower court’s judgment and held that Google’s copying of the code was not fair use and that the copyrightability issue that was resolved in its 2014 decision still applied. In 2019, Google filed its petition for Supreme Court review once again, and finally the time for oral argument arrived.

The main issues for the Court to decide are a review of the Federal Circuit decisions—whether Oracle’s code was copyrightable to begin with and whether Google made fair use of the code. But also, and perhaps most importantly, the Court must decide what standard of review should be used to make a fair use determination. This last question will have a lasting impact far beyond this particular dispute.


Under the Section 102 of the Copyright Act, software is considered a “literary work” and is afforded legal protection. Oracle would therefore have the right to prevent others from using its code or, as was the case, license the use of its software for others to use. But Google makes an argument under the merger doctrine that the code was never subject to copyright protection. The merger doctrine states that when an idea and its expression are inextricably merged there cannot be any copyright protection for the idea. This doctrine gets right to the heart of copyright law: original expression is protectable; an idea—a method, a system, a way to explain something—is not. Google’s argument is that there is only one way to express Oracle’s code, and therefore the code itself cannot be copyrightable at all. Google repeatedly suggests that to rule otherwise would be imparting protection over a method, which is something reserved for patent law.

The Court, and perhaps the rest of us not well-versed in computer science, struggled with this concept in the context of a very technical set of facts. Google’s counsel, Thomas Goldstein, drew comparisons to the Baker v. Selden Supreme Court case where Selden created a new method of bookkeeping and attempted to assert copyright protection over his ledger forms. Goldstein likened Oracle’s code to those ledgers—innovative and hugely successful, but merely a methodology that does not warrant copyright protection. The justices took turns proposing different hypotheticals to challenge this argument. First was Chief Justice Roberts, “cracking the safe may be the only way to get the money that you want, but that doesn’t mean you can do it,” followed by Justice Thomas’s analogy to a football team’s playbook that Google swiped for its own win. Mr. Goldstein argued back that copyrighting the knowledge itself of how to crack a safe would not be allowed and that Google is not “trying to take someone’s fan base or their football players or anything else.” The justices did not seem convinced by Google’s assertion that Oracle’s method of writing code was the only way to create the function Google wanted for its Android platform.

It is easy to get lost in the technical language of the case, between declaring code and implementing code and shortcut programs. Oracle’s counsel, Joshua Rosenkranz, cut through all those distinctions saying, “code is code,” and he pointed out that no circuit court drew a distinction between the types of code. Congress could have carved out different types of code in its consideration of the Copyright Act, but it expressly chose otherwise. Mr. Rosenkranz also circled back to distinguish Oracle’s code from the accounting ledgers in the Baker v. Selden case, highlighting that Oracle is not trying to prevent others from creating their own programs. “Others are free to write and organize their own prewritten programs however they see fit, as long as they don’t copy ours.” Mr. Rosenkranz underscored, and multiple justices acknowledged, that other competitors like Apple and Microsoft did not resort to pure copying to create their products, and that companies like IBM and SAP paid Oracle to properly license the code. Just because Oracle’s method was the most elegant, and therefore popular, way of expressing the code does not make it the only method of doing so. Mr. Rosenkranz summarized this issue well, saying, “the Copyright Act does not give Google a pass just because it would be expensive to recreate our expression.” 

Fair Use

Even if the code were copyrightable in the first place, Google argues, fair use still justifies its copying of the code. Fair use is an affirmative defense to an accusation of copyright infringement. A person or entity who has plainly copied another’s protected work can argue fair use to justify their taking and subsequent use of the copyrighted material. Congress has enumerated a four-factor test within the Copyright Act for courts to follow when evaluating a fair use claim—a test that the Supreme Court last used 26 years ago in Campbell v. Acuff-Rose Music—weighing the purpose and character of the use, the nature of the original work, the amount and substantiality of the portion used, and the effect of the use on the market value of the original work.

Modern courts, when evaluating the first factor of purpose and character of the infringing use, often focus on whether the use was transformative. Google has used this argument successfully in the past, in Authors Guild v. Google, defending its copying of millions of books for use in its Google Books search engine. The court in that case found that digitizing books for use in an online search database was highly transformative. In this case, Mr. Goldstein tried to make a similar argument that Google wrote an improved version of Oracle’s Java code to be more suitable for use in a smartphone. Mr. Goldstein said that this is “what Congress would want, that is to be able to take the functionality of a computer program, [and] someone else comes along and does it better.” He then stated that computer code only performs one function and therefore can still be considered transformative even if it does nothing different than the original work, only to later argue that the word “transformative” is not used in the Copyright Act and might not even be the correct standard of evaluation in this case.

Google also maintained a policy argument that “reusing software interfaces is critical to modern interoperable computer software,” and that to rule in Oracle’s favor would “upend that world” and make computer programming so inefficient as to have fewer creative computer programs. Justice Kavanaugh pointed out that despite Google and its amici claiming a dire result if the Court should rule for Oracle, in the years since the Supreme Court first denied certiorari, “I’m not aware that the sky has fallen.”

Oracle made a more direct argument based on legal precedent, noting that the Supreme Court had already held in Harper & Row and in Stewart v. Abend that no court had found or upheld a fair use verdict where an infringer copied so much of an original work for use in a competing commercial product, all while retaining the same meaning and purpose as the original work. Mr. Rosenkranz pointed out that Google conceded that every line of code copied serves the same purpose in Google’s Android platform as it did in the original software—there was no alteration or transformation of the original code—and Malcolm Stewart, arguing on behalf of the United States in support of Oracle, echoed that point. Mr. Rosenkranz then cited the Court’s own decision in Campbell, saying that what Google did is the “epitome of commercial superseding use: using a work ‘to get attention or to avoid the drudgery in working up something fresh.’”

Standard of Review

The discussion quickly turned to the standard of review that the Court should use to evaluate fair use—whether the determination must be left up to a jury, like the jury that ruled in Google’s favor previously, or whether it can be made by a court, such as the Federal Circuit that overturned the jury verdict to side with Oracle. Google argued that fair use is determined by “whether the jury could reasonably find fair use,” and that “no previous court ever held that only a court may decide fair use.” Mr. Goldstein argued that giving the decision to a jury is the better choice due to the numerous factors involved in a fair use verdict and that jury instructions exist to give appropriate legal certainty and guidance to the jurors.

Oracle disagreed and argued that de novo, meaning “anew,” is the correct standard to resolve this question. This would allow for a court to make a fair use determination without being bound by the decision of a previous court or verdict. Mr. Rosenkranz again cited Harper & Row: “an appellate court may conclude, as a matter of law, that the challenged use does not qualify as fair use once it has the factual record and resolves all subsidiary factual questions in favor of the fact-finder.” Judicial review provides certainty to a fair use determination that would be lost if left to a lay jury. Mr. Rosenkranz also noted that making fair use a fact-finding mission that only a jury may evaluate would all but preclude summary judgment determinations, which are exclusively judge-made decisions and currently comprise the vast majority of fair use cases.


The stakes in this case are extremely high—if it was not made clear by the number of times the justices and attorneys referenced the sky falling—not only for Oracle and Google and the computer science world at large, but also for any copyright creator or user who will eventually face a fair use determination. Joshua Simmons, co-counsel for Oracle and partner at Kirkland & Ellis, looks forward to the Court’s decision, writing:

Given that Google’s copyrightability argument would make it difficult for any code to be protectable, I am hopeful that the Court will follow the text of the Copyright Act and the intent of Congress that code, like Oracle’s, is protectable. Likewise, I hope that the Court will follow a century of precedent allowing courts to decide as a matter of law that, when the defendant creates a substitute that competes with the original, it simply is not fair use.

People of all industries, from professionals to hobbyists, should take note of last week’s oral argument and what will be an historic decision from the Court.