Editor’s Note: Juristat doesn’t take a position in the software patentability debate. The views reflected in this piece are meant to be informative and do not necessarily reflect the views of Juristat or the authors themselves.
For most of the USPTO’s history, technological innovation has been mainly in the form of mechanical innovation—a better leaf blower, a more efficient engine, a deadlier mousetrap, etc. All of this changed during the second half of the twentieth century, which saw a seismic shift from mechanical innovation to technological innovation through hardware and software.
Juristat has kept a close on eye on recent trends in the software patents space, especially the effects of the Alice decision. We have written numerous articles and hosted several webinars explaining the case’s impact on the software industry with data, and even developed an analytical tool that patent prosecutors can use to help keep their applications out of Alice-heavy art units. At this point, we thought it might be a good idea to take a step back from in-depth statistical analysis to cover some of the broader arguments driving the conversation about the patentability of software.
In this article, two of our attorneys and marketing specialists, Austin Underhill and James Cosgrove, will each take a side in the debate and present several of the most common arguments as to why software should or should not be patentable.
by Austin Underhill, JD
Under Section 101 of the Patent Act, one may receive patent protection for any process, machine, manufacture, or composition of matter. Precedent provides three exceptions to these generally accepted subject matter: laws of nature, physical phenomena, and abstract ideas. Software falls into the last category.
Software code itself is merely a form of language with no physical embodiment. Intellectual Ventures I LLC v. Symantec Corp. (Fed. Cir. 2016) (Mayer, J., dissenting). It is a set of instructions meant to be executed by a computer. “Because generically-implemented software is an ‘idea’ insufficiently linked to any defining physical structure other than a standard computer, it is a precursor to technology rather than technology itself." Id. It is simply an idea written in a unique language, and ideas alone are not patentable.
Some will argue that software is more than code; that it’s the user interface and user experience (UI and UX, respectively) itself. Perhaps there’s even some truth to that argument. However, the modern trend toward UI-based design patents suggests the USPTO feels differently. After all, if utility patents encompass UI, why then are companies like Apple seeking design patents on the very same thing? In fact, according to Tech Crunch, “the fastest growing segment of design patent filings is in the UX/UI space.”
If we accept this distinction between code and UI, it becomes clear that code is merely the expression of an idea that manifests itself in a UI. So maybe it comes down to semantics and how we define software. Code, as the expression of an idea, should not be patentable. The resulting UI is, perhaps, a different story, but we’ll save that discussion for a different article.
Patent ineligibility doesn’t leave software without protection; all we have to do it is look elsewhere in the intellectual property landscape. “Copyright, a form of intellectual property law, protects original works of authorship including literary, dramatic, musical, and artistic works, such as poetry, novels, movies, songs, computer software, and architecture.” (What Does Copyright Protect? [emphasis added]).
This quote from the Electronic Frontier Foundations amicus brief in Alice is instructive:
The software market began its rapid increase in the early 1980s—around the time personal computers became widespread—more than a decade before the Federal Circuit concocted widespread software patents in 1994. . . . Obviously, no patents were needed for software to become a $60 billion/year industry by 1994.
Put another way, the software market didn’t just survive without patent protection, it thrived. In fact, IBM argued in a 1972 Supreme Court case that “Copyright achieves the basic purpose of preventing one who has made no investment of time or money in creating a computer program from appropriating the program by copying it.” To be fair, IBM has since changed its mind, but I suspect that has more to do with licensing profits than some new revelation about the inadequacies of copyright law.
The patent system, as it pertains to software development, is no longer a defense against infringement but rather a weapon in the arsenal of non-practicing entities (NPEs) and large corporations. In fact, it’s the increase in software patents that has caused NPEs to grow in number and power. (Defend Innovation: How to Fix Our Broken Patent System). Unsurprisingly, NPEs, also known as “patent trolls,” initiate the majority of patent litigation. Id. While giant corporations can afford to defend themselves, startups and inventors are out of luck. Clearly, the system protects the big guys and eliminates perceived competition. Id. With all this litigation, companies are forced to play defense by stockpiling broad software patents. Id.
Most software patents are intentionally vague descriptions of an idea without any source code. Intellectual Ventures I LLC v. Symantec Corp. (Fed. Cir. 2016) (Mayer, J., dissenting). All these vague and overbroad claims create patent thickets - “a dense web of overlapping intellectual property rights that a company must hack its way through in order to actually commercialize new technology.” (Navigating the Patent Thicket: Cross Licenses, Patent Pools, and Standard-Setting). In this way, the system actually hinders innovation. Id. Put another way, these filings do little to promote public disclosure, and these consequences are antithetical to the objectives of our patent system.
by James Cosgrove, JD
One of the most common attacks on the patentability of software is that software is simply a series of algorithms based on mathematics, and that mathematics is an unpatentable abstract idea. While it is true that software patents rely a great deal on mathematical algorithms, this is the case for almost all patented inventions. Mathematics is the language of the universe, and all utility patents can be reduced to mathematics at their most basic level. (Mossoff; A Brief History of Software Patents (And Why They’re Valid).
Prohibiting software patents because they are based on mathematical algorithms would essentially invalidate almost every utility patent. The computer code that software runs can more accurately be conceived of as a method of expression like the English language. Thus, prohibiting software patents because they are based on mathematics would be the equivalent of prohibiting other utility patents because they are written in English. Id.
In the early days of computer programming, there was no “hardware” and “software;” all computer programs ran on machines dedicated solely to that one program. During the PC Revolution of the 1980s led by Steve Jobs and Bill Gates, among others, hardware and software began to be severed into separate products. Instead of purchasing a word processor and a calculator, for example, consumers could now purchase a computer with a central processing unit (CPU) capable of running both a word processor and a spreadsheet program, each of which is based on a series of instructions the software program gives to the CPU. Thus, the primary value of software is the functionality it delivers to the end user. Id.
Copyright law protects words alone and not the broader idea or function represented by those words. Patent law, by contrast, protects ideas and functions, but not the specific words that describe those functions. The problem of limiting software to copyright protection is that any enterprising software developer can simply tweak the code of a particular program while maintaining its core functionality without running afoul of copyright law. Id. Because the value of software is in its functionality rather than in its method of expression, patent law is the only appropriate protection for it. Id.
No one doubts that computer hardware is patentable. A piece of computer hardware is a perfect example of what courts consider to be a “specific machine,” which is an important clue to patentability. Bilski v. Kappos, 561. U.S. 593 (2010). Obtaining a patent on software is much more controversial because software is not tied to a specific machine in the classic conception of the term, but is instead meant to be run on any general purpose computer. However, software’s adaptability does not negate the fact that software and hardware are virtually interchangeable. (Goetz; Should Patents be Awarded to Software?)
It is well-known in the technological arts that the functions of a particular program can be either wired into a chip as hardware or coded as software. Id. The only difference is that a software program costs less to produce and is more efficiently sold to end-users than hardware. Id. In fact, the Software Alliance argued in an amicus curiae brief to the Alice court that software actually has the capability to turn general-purpose machines into specific machines, albeit for a short period of time:
With the stroke of a finger, a tablet computer or smartphone can instantly become a GPS navigation system, a word processor, a camera, a video or music player, an eReader, or an electronic piano. Thus, while the computer itself consists of a fixed set of standard physical components, it can be given almost limitless functionality through software that reconfigures the electronic pathways running through millions of transistors embodied in the computer’s processor, literally (if temporarily) converting the computer into a special purpose device.
To illustrate this concept with an example, the invention of a word processing program during the PC Revolution of the late twentieth century is the functional equivalent of the invention of a typewriter during the Industrial Revolution. (Mosoff). In the intervening years, our use of technology has shifted away from owning many machines that each perform one task to owning one or two machines that perform an infinite number of tasks. The only difference between these two technological paradigms is that the functionality the consumer seeks today is sold on the marketplace as software instead of as single-function hardware. Id.
If the way we interact with our machines has changed, why should patent law not respond?
This article is merely a summary of the arguments that attempts to give an overview of the many issues surrounding software patentability. While the topics herein could easily fill a law review article, suffice it to say that the software patent debate is not likely to end any time soon. If one thing is clear, however, it is that patents on software are becoming increasingly difficult to obtain in the post-Alice legal landscape, as we have noted here, here, and here. Luckily, there are tools available to attorneys working with these technologies to help them avoid the most burdensome effects of recent § 101 jurisprudence.
Take the uncertainty out of prosecuting software applications with Juristat Drafting, a groundbreaking data analytics tool allows prosecutors to draft their claims in such a way to as to avoid art units where they are certain to receive an Alice rejection. See how it works, set up a demo today.