Broad Support for Google in the First Round of Supreme Court Briefing
Last week, Google filed its opening brief requesting that the Supreme Court reverse the decisions of the U.S. Court of Appeals for the Federal Circuit (“CAFC”) in Oracle’s long-running copyright infringement litigation against Google. (See here for a summary of the litigation until this point.) This week, 26 amicus briefs have been filed in support of Google, and 2 in support of neither party. These briefs are digested below, relying on the briefs’ summary of argument.
Google’s Opening Brief
Google’s brief first argued that Oracle had no copyright interest in the declarations of the Java SE libraries under the merger doctrine. Google reused declarations from the Java SE libraries because no other option would recognize the calls used by Java developers. And because no other instructions can perform the declarations’ function, merger excludes them from copyright protection. Any other result would impermissibly convert Oracle’s copyright in the declarations’ expression into an exclusive right to the declarations’ function. This is true whether the declarations are considered individually or collectively. Oracle’s claim would upend the well-settled understanding that the functions of earlier computer software may be reimplemented, including by reusing instructions required to replicate the commands known to the earlier product’s users. Reimplementation encourages developers to create valuable new computer software that improves on legacy products. Oracle’s copyright theory would give software companies the power to block reimplementation and to inhibit the development of new and better software.
Google then argued that even if the Court ruled that the declarations fell within the scope of copyright protection, the CAFC erred by overturning the jury’s fair use verdict. The jury was entitled to conclude that Google’s reuse of the declarations was consistent with the overarching purpose of the fair use doctrine: avoiding the rigid application of copyright that would stifle creativity. Additionally, there was sufficient evidence to support the verdict with respect to the jury’s weighing of each of the four fair use factors.
The amicus briefs can be divided into four groups: industry briefs; law professor briefs; public interest briefs; and cautionary briefs. All the briefs can be found on the Supreme Court’s docket page for the case.
• The amicus brief submitted by the Computer & Communications Industry Association (“CCIA”) and Internet Association (“IA”) provided historical context for understanding the dispute and the adverse impact affirmance of the CAFC’s decision would have on innovation. (I was the counsel of record on this brief.) The history of the technology industries has been defined by the evolution of copyright rules governing interoperability. Without understanding the development of the rules that enable today’s interoperable world, one could not fully appreciate how severely the CAFC’s decisions contradict decades of industry norms. The brief described how a series of judicial and legislative decisions around the world laid a foundation of interoperability that modern technology industries have relied upon for decades to compete and innovate.
• In its brief, Microsoft stated that it had a unique—and balanced—perspective on the technological, legal, and economic issues this case implicates. On the one hand, Microsoft relies on copyright protection, which enables it to license its own products and services and earn a fair return for its creations. On the other hand, Microsoft recognizes that limits on copyright—such as the fair use doctrine—are essential to the operation of the computer industry. Microsoft criticized the CAFC’s unduly narrow test for transformative use of functional works such as software. Microsoft argued that using declarations for the same purpose in a different platform could be transformative. The CAFC’s overly-restrictive application of fair use would harm collaborative efforts in the software industry, which rely on the breathing room provided by fair use.
• IBM and Red Hat asserted that software interfaces are crucial to every critical technology that interacts with the cloud to expand and enrich users’ experience with connected devices, from self-driving cars to personal fitness apps. It has long been understood that software interfaces, as distinct from the software implementations of those interfaces, are not copyrightable subject matter. Properly understood, software interfaces embody systems and methods of operation that § 102(b) expressly excludes from copyright protection. As support, the brief cited Perris v. Hexamer, 99 U.S. 674 (1879), which found that the copyright in maps did not extend to a “system of coloring and signs” for identifying real property characteristics or to a “key” which explained symbolic meanings of coloring and signs. Unrestricted use of software interfaces is an essential aspect of all software development and has long promoted competition, widespread innovation, and progress in the computer, information technology, communication technology, and networking fields. Innovation in these fields would be impeded, not advanced, if software interfaces were now deemed outside of 17 U.S.C. § 102(b) and, as such, a basis for opportunistic attempts to control functional bridges between independently developed software products.
• Mozilla, Medium, Reddit, Creative Commons, Etsy, Wikimedia, and other technology organizations stated that for decades, software engineers have relied heavily on reimplementation, including reuse of functional protocols such as the software interfaces in this case, to create competing alternatives to incumbent industry players and develop new markets without fear of copyright infringement. In accord with this Court’s ruling in Baker v. Selden, 101 U.S. 99, 105 (1879), and the plain language of 17 U.S.C. § 102(b), the software industry has flourished utilizing this approach to make internet and software solutions more accessible, affordable, diverse, and robust. By reversing this rule in the context of API packages, the Federal Circuit upended decades of industry practice and the well-established expectations of developers, investors, and consumers.
• The Retail Litigation Center, which represents the U.S. retail industry, stated that this dispute is but one example of a more global development in the retail industry: the increasing ability of incumbent technology platform providers to use their market power to exploit their positions, foster inefficient “lock-ins,” and thus further entrench their market power. The CAFC’s decision magnifies the gravitational pull of these developments by imposing new hurdles that make it even more difficult for retailers and other businesses to switch to new software systems, even when those systems are better. Allowing reimplementation is a key measure to prevent lock-in, check incumbent power, and ensure competition, not merely in the computer marketplace but also in the retail industry. Affirmance of the CAFC’s decision would enable incumbents to weaponize copyright law in an anticompetitive fashion.
• The Developers’ Alliance argued that the Court should recognize that use and reuse of interoperable software is the foundation of innovation in the software development industry. Subjecting API declarations to copyright protections would force developers to constantly re-engineer something that already works, stymying creativity and innovation.
• The Python Software Foundation, Tidelift, Open UK, and Protocol Labs argued that the Court should apply its “conceptual separability” test from Star Athletica, L.L.C. v. Varsity Brands, Inc., 137 S. Ct. 1002 (2017). Under this test, the expressive elements of a mixed work must be conceptually removed from the functional elements of the work. Only if there are expressive elements that are separable from the functional elements of the useful article does the court proceed to evaluate infringement. Further, the CAFC erred by overlooking Oracle’s conduct in the marketplace when assessing fair use—in particular, Oracle’s decision to release its software under a GNU General Public License. It encouraged other developers to use its APIs, which is exactly what Google did.
• 83 computer scientists represented by the IP and Innovation Clinic at Stanford Law School filed a brief stating that the CAFC’s decisions undermined a fundamental process—software interface reimplementation—that has enabled innovation in the software industry for decades. They asserted that software interface reimplementation is a long-standing, ubiquitous practice that has been essential to realizing fundamental advances in computing. The decisions also reflected a misunderstanding of the difference between programs and interfaces.
• Software and Systems Developers for U.S. Government Systems expressed deep concern that the CAFC’s decisions hand some copyright holders the ability to monopolize systems, processes, and methods of operation of others merely because those systems, processes, and methods interact with or otherwise utilize an interface embodied in the copyright holder’s product. The result of the CAFC’s unwarranted expansion of copyright law, violating the boundary between copyright and patent, will be that technology and communications infrastructures, systems, and services will become more fragmented, less standardized, and less interoperable, all to the detriment of technical progress and efficiency, and of the Progress of Science and useful Arts.
• Engine Advocacy argued that this lawsuit represents an end-run around the carefully constructed requirements and limitations of patent law. Oracle seeks to use copyright to obtain a monopoly on a functional system—its API declarations. But protection of ideas, systems, and methods is and has always been the domain of patents.
• Two Empirical Legal Researchers who are also programmers, James Daily and Adam Feldman, wrote that software developers would be hamstrung if they were limited to a universe of language based on words and phrases not used in the past. Requiring every computer programmer to reinvent common phrases and concepts would make new works difficult to write and confusing to read. Likewise, it would prevent the cumulative development of new ideas and concepts that is vital to innovation in computer science, just as it is in the law.
• Software Innovators, Startups, and Investors argued that the CAFC’s rulings in this litigation have changed the rules of copyright. By holding that the software interfaces at issue here, which allow developers to access prewritten lines of code, are copyrightable, and use of them will rarely, if ever, constitute fair use, the CAFC struck a blow against the interoperability copyright meant to protect—a blow that falls particularly heavily on startup companies. That is because startups have been able to thrive largely because they can use software interfaces like Java’s to create new and innovative products that are interoperable—interacting seamlessly with networks, hardware, and software. The settled expectation of the startup community is that these interfaces will be readily available.
• Auto Care Association and Static Control Components noted that virtually every modern product is controlled by software. To diagnose and repair these products, and to create competitive and improved products, a wide range of businesses depend on the right to access and interface with software. The CAFC’s decision threatens that right to repair by granting overbroad copyright protection to the declaring code of APIs.
• Rimini, a third-party support provider for Oracle software products, described how affirmance would allow Oracle to stifle competition in the software service market.
Law Professor Briefs
• Professor Pam Samuelson filed a brief on behalf of 72 intellectual property law scholars. This brief argued that until the CAFC’s 2014 decision, software developers felt free to compete and innovate in the development of compatible software because major decisions from the Courts of Appeals for the Second and Ninth Circuits had established that copyright law does not protect software interfaces that enable the development of compatible programs. It adopted an unduly narrow view of this Court’s ruling in Baker v. Selden, 101 U.S. 99 (1880), which excluded methods, systems, and their constituent elements from copyright’s scope. It ignored Congress’ codification of the method/system exclusions. It misconstrued the case law properly interpreting those exclusions in relation to program interfaces. The CAFC also misapplied the merger doctrine and case law persuasively holding that interfaces that enable compatibility are unprotectable by copyright law. Programmers should have to write their own implementation code, as Google did, but interfaces that enable compatibility should be free from copyright restrictions.
• Professor Rebecca Tushnet filed a brief on behalf of seventeen law professors focusing on fair use. The CAFC’s rejection of a jury finding of fair use embraced a rigid approach that, as a matter of law, would bar any copying of code into a new program, even for a different platform, as non-transformative and unfair. The CAFC made key mistakes about the fair use factors and their balancing, including inflating the relevance of commerciality, applying an erroneous “no more than necessary” standard for copying, dismissing as insignificant the highly functional nature of computer programs, and conflating the market for Java SE as a whole with the market for individual declarations. Absent these legal errors, it is clear that the jury was at least reasonable in making factual findings that supported a finding of fair use.
• Professors Peter Menell, David Nimmer and Shyamkrishna Balganeshade argued that the CAFC’s decisions conflict with the Supreme Court’s ruling in Baker v. Selden, misinterpret Congress’s codification of the idea/expression dichotomy in 17 U.S.C. § 102(b), and upend nearly three decades of sound and well-settled decisions on the scope of copyright protection for computer software.
• Three Civil Procedure and Legal History Professors argued that the CAFC wrongly applied de novo review to overturn the jury verdict of fair use. Doing so was unprecedented and unconstitutional.
• Professor Glynn Lunney offered the Court a summary of recent empirical research on whether more incentives lead to more creative output. The data show that more incentives do not lead to more creative activity. This suggests that providing additional incentives to Oracle by ruling in its favor would not lead to increased creative output of original works, such as Java. In contrast, there is no similar reason to doubt that a ruling in favor of Google would increase follow-on authorship. The Court should therefore rule in favor of Google.
• Professor Michael Risch argued that the CAFC improperly examined the copyrightability of the Java declaration in the abstract, as opposed to in the context of an infringement analysis.
Public Interest Briefs
• The Center for Democracy and Technology, the Institute for Intellectual Property & Social Justice, National Consumers League, and National Federation of the Blind argued that consumers benefit from the freedom to use and reuse interfaces, and would pay the cost if that freedom were taken away. If copyright law were expanded to protect interfaces, fragmentation, higher prices, and frustration would result. Fewer programs and devices would be able to talk to each other. Many efforts to make technology accessible to the blind would be thwarted. Standard interface design would become a rarity, making each new program more burdensome to learn and making each program its own separate silo.
• In its amicus brief, the Electronic Frontier Foundation argued that the CAFC improperly rewrote the language of § 102(b) by effectively deleting seven of the statute’s eight exclusions of copyrightable subject matter. The CAFC incorrectly limited the section to the first exclusion, which prohibits copyrighting an “idea.”
• The Software Freedom Law Center argued that in overturning the jury verdict on the issue of fair use, the CAFC flagrantly ignored the teachings of this Court on the appropriate standard of review under Fed. R. Civ. Proc. 50. The CAFC insisted on reweighing the evidence, refused to draw inferences favorable to the non-moving party and adopted inferences favorable to the moving party. The CAFC based its own inferences on factual misunderstandings concerning contemporary software copyright licensing practices.
• The amicus brief of the American Antitrust Institute observed that the CAFC’s rulings failed to consider that copyright law seeks to promote innovation and consumer welfare by preserving a balance between exclusive rights and competition. By ignoring the competition side of the equation—indeed, by rejecting the relevance of interoperability considerations—the CAFC’s rulings, if not overturned, may slow innovation and competition in software-dependent markets.
• R Street Institute, Public Knowledge, and the Niskanen Center argued that the reimplementation of interfaces is prevalent, expected, and valuable. Computer command languages are commonly reimplemented. Cloud computing relies extensively on APIs, and reimplementation of leading firms’ cloud computing APIs is frequent practice for competitors, including Oracle itself. Internet communications and computer peripherals similarly depend on reimplementation, by virtue of their conformance to technical communication standards that incorporate APIs. Governments are also involved in the business of API reimplementation. Federal agencies and other government entities produce, reimplement, and direct reimplementation of APIs to serve public objectives such as health care interoperability, aviation safety, and law enforcement coordination.
• The American Intellectual Property Law Association filed a brief in support of neither party, but the positions it advocated clearly favor Google. It wrote that the declaring code portions of the Java APIs are methods of operation to which copyright protection does not extend, while the implementing code of the Java APIs are not, and thus are subject to copyright protection. The Java API declaring code is the essential and only means by which programmers can operate and access the implementing code of the Java API. The relevant portions of the declaring code comprise merely the name of the function being called or the names of the inputs and outputs. The implementing code, by contrast, provides the particular expression of how each command is executed. Consequently, the declaring code is a non-copyrightable method of operation, while the implementing code contains protectable expression. AIPLA stated that it is vital that the Court maintain the fair use exception to copyright infringement as a flexible, case-by-case balancing of the fairness of each use. The Court should abrogate all lower court rulings that give any one fair use factor greater or lesser weight in all cases. Where a jury renders a general verdict on fair use, that verdict should be given deference, and any implicit factual findings may be overturned only if these findings are unsupported by substantial evidence. Until the decision below, no Court of Appeals had exercised de novo review to reverse a jury’s general verdict finding of fair use. Indeed, over the last two centuries, courts have treated fair use as a factual question within the province of the jury to resolve.
The AIPLA brief might be the most surprising one filed. AIPLA represents the patent bar, and typically supports the position of rightsholders. Its balanced position here perhaps reflects a concern that the CAFC used copyright to grant patent-like protection, thereby threatening the equilibrium in the patent system.
Finally, two briefs urged the Court to ensure that its fair use ruling concerning the Java SE declarations did not affect other fair uses.
• The American Library Association, the Association of College and Research Libraries, the Association of Research Libraries, and Software Preservation Network noted that the Court’s decision in this case could affect far more than the reuse of software interfaces. The transformative use jurisprudence that has evolved in the quarter century since Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994), has enabled a wide range of technology-dependent uses, ranging from search engines to user-generated content. This new understanding of fair use is particularly critical to research, education, and librarianship in the Twenty-First Century. If the Court takes up the issue of whether Google’s use in this case was fair, the associations urged the Court to exercise great care to ensure that its reasoning in no way jeopardizes amici’s transformative uses in support of research, teaching, and learning. The brief suggested that the safest way to avoid disrupting the fair use ecosystem is to avoid the issue altogether by deciding the case in Google’s favor on the basis of merger.
• The Robert Rauschenberg Foundation and the Andy Warhol Foundation, in a brief supporting neither party, stated that whatever the Court chooses to say about fair use in the software context, it should be careful not to suggest inadvertently that the same analysis necessarily applies to visual art, or in other artistic contexts such as literature or music.
The Legacy of ACIS
Four of the briefs mentioned the American Committee for Interoperable Systems (“ACIS”). The CCIA/IA brief explained that the pro-interoperability judicial decisions issued in the 1990s did not arise in a vacuum. Rather, Sun Microsystems, the company that wrote the Java SE libraries and thus the software interfaces at issue in this case, convened a meeting in December 1991 at its headquarters in Silicon Valley to discuss troubling case law that threatened interoperability. The companies at this meeting formed ACIS, and selected Sun’s Deputy General Counsel Peter Choy as its chairman. ACIS agreed upon a Statement of Principles declaring that “[t]he rules or specifications according to which data must be organized in order to communicate with another program or computer, i.e., interfaces and access protocols, are not protectable expression under copyright law.” ACIS subsequently filed amicus briefs in support of interoperability in many of the seminal software copyright cases during the 1990s, with Peter Choy as counsel of record. Oracle endorsed the ACIS principles when it joined ACIS. And both Oracle and Sun were members of CCIA, which joined ACIS in some of these briefs.
The Samuelson brief noted that ACIS filed an amicus brief in the Supreme Court in Lotus v. Borland arguing that a software interface was a system or method of operation outside the scope of copyright protection. The Tushnet brief quoted this brief at length in support of the argument the Google had a good faith belief that its reuse of the Java declarations (written by Sun) was lawful. This good faith weighed in favor of a fair use finding. IBM/Red Hat cited a different ACIS brief to refute Oracle’s argument that “software interface” was a term Google invented for its cert. petition.
The CAFC ignored this history of Sun and Oracle supporting the positions Oracle now opposes so vehemently. The Supreme Court, however, may be more interested in this complete reversal.
Oracle will have the opportunity to respond to Google and its amici on February 12. Oracle’s amici can file briefs by February 19. Google’s reply brief is due March 13. The oral argument likely will take place in late March.