More than two thirds of all patents challenged under 35 U.S.C. §101 have been invalidated since Alice Corp v. CLS Bank was decided in 2014.[1] Is this recent trend signaling the beginning of the end of the software patent? Should software even be patentable? Will a clear test help? While both the majority and dissenting opinion in Intellectual Ventures I LLC v. Symantec Corp. agree that software is patentable, in a bizarre twist, the concurring opinion has declared that software patents are finished. Clearly some judges on the Federal Circuit have run out of patience with the multitude of software patents that were drafted prior to the Mayo/Alice cases being decided. Nonetheless, don’t expect a clear test for patent eligibility under §101 anytime soon. This article reviews the majority and dissenting opinions in Intellectual Ventures, contrasts the concurring opinion, and explains why we believe a clear test for patenting software is not needed and in fact, would set back the patent system for years.
Intellectual Ventures I LLC v. Symantec Corp.
Intellectual Ventures sued Symantec for infringement of three patents. Ultimately, all three were found to be directed toward different abstract ideas:
1) Receiving mail and discarding it based on the characteristics of the mail
2) Screening messages
3) Virus scanning
The Federal Circuit went through both steps in the Alice framework and ruled that all three patents were invalid under §101 (affirming the District Court on two and overruling the Court on the remaining patent).
The Federal Circuit made two things clear. First, the inventive concept required to transform an abstract idea into a patent eligible concept must be in the claims. This concept was spelled out by the majority directly addressing the dissenting opinion. The dissent had argued that one claim in one patent was patentable because that claim improved the functioning of the computer and addressed problems specific to the internet. However, the majority stated that, while it was true the patent disclosed an improvement in the functioning of a computer, the improvements at issue were absent from the claims. Therefore, the Federal Circuit held the claim invalid.
Second, the majority makes clear that software is still patent eligible. The majority restated precedent noting that to be patent eligible, software must improve the functioning of the computer or solve problems specific to the technological environment. The majority even gave an example of how the virus screening claim at issue might have been patent eligible.[2] The fact that the majority stated what they are looking for when determining software patent eligibility and provided a concrete example of how such a claim might have been patentable, makes clear that software is still patent eligible.
Judge Mayer’s Concurring Opinion
Initially, Intellectual Ventures I LLC v. Symantec Corp. seemed like another run-of-the-mill software patent case. Company A sues Company B for infringement of software patents. Company B argues that the asserted patents are invalid under §101. The Federal Circuit agrees and the software patents are ruled invalid. Case over, right? Not so fast. Judge Mayer, in a concurring opinion, has decided he’s had enough of software patents in general. His frustration likely built up after more than two years of purging the system of software patents that never should have been issued. Since Alice in 2014, software patents have been invalidated at the Federal Circuit level under §101 at an alarming rate of roughly 95 percent.[3]
Judge Mayer’s central point, on its face, is difficult to dispute. If an idea (software) is not patentable and only embodiments of the idea are patentable, and the generic computer the software is running on is not patentable, then all ideas running on the generic computer should not be patentable. However, guidance from the Courts, like the majority opinion, has said software must improve the functioning of the computer or solve problems specific to the technological environment in order to be patent eligible. It’s undisputable that patents directed at conventional ideas cannot be patented by simply tying those claims to a generic computer. What we believe Judge Mayer is missing is that not all software patents are generic ideas on generic computers. In reality, a lot of software patents are behind the improvements of the electronic devices we use today. Software has a place in patent law; unfortunately, it has taken patent law several years to catch up and find that place.
Judge Mayer’s reasoning has two main points:
1) Software patents “run afoul” of the First Amendment
2) Software patents on a generic computer are not eligible for patent protection
Judge Mayer’s first point regards preemption, a main concern in the post-Alice world. However, instead of worrying about how a patent claim might preempt a field of invention, Judge Mayer expresses concern about preempting the First Amendment by “exacting heavy taxes on widely-used conduits for online expression.”[4] This concern, while somewhat valid, is actually resolved by the Alice framework, which specifically addresses the potential for preemption. If, for example, an idea preempts “widely-used conduits for online expression,” it would be ineligible for patent protection under §101. Thus, Judge Mayer’s slippery slope argument involving preemption is not a valid reason to make software ineligible for patent protection.
Judge Mayer’s second point, the more sweeping concept of preventing patents from being issued on software, is broken into four sub-points:
1) The scope of software patents outweighs their technical disclosure
2) Software patents provide incentives at the wrong time
3) There are too many software patents
4) Software patents lack the definiteness required by patent law
The first sub-point also regards preemption. As noted above, preemption is accounted for under the current Alice framework. However, one sticking point for Judge Mayer is that most software patents do not include the software code behind the invention. The reason for the lack of code in the patent, however, is that the code itself is not patentable. What is patentable is what the code does. Software code itself can be protected using copyright law and has no place in patent law.
The second sub-point, that software patents provide incentives at the wrong time, exists for virtually any invention, not just software. While Judge Mayer correctly points out that a lot of software patents are filed at the “idea stage,” before the invention is finished, the same is true for most inventions. This problem has only gotten worse because of the new First to File rule under the America Invents Act. It’s true that “those who scamper to the PTO early…reap hefty financial dividends.”[5] But, this reward is not a result of software patents; it is a result of the new filing provision of the America Invents Act. Right or wrong, first-to-file is here to stay and all inventors are incentivized to file patent applications as early as possible.
The third sub-point, that there are too many software patents, should have no bearing on whether software is patent-eligible. Clearly, most of the things we use today are operational because of software. In fact, it is very likely that you are reading this article using an electronic device that is operational because of software. It’s no surprise that the most popular area of innovation has a lot of patents. Software’s patent eligibility doesn’t hinge on the popularity of the technology it relates to.
The fourth sub-point, that software patents lack the definiteness required by patent law, is also related to preemption. Judge Mayer states that software is “akin to…literature or a piece of music, undeniably important, but too unbound” to be patent eligible.[6] But Judge Mayer misses the point – software patents don’t patent software, they patent what software does. If software simply does something that can be accomplished without it, the Alice framework will render that ineligible for patenting, thereby preventing the preemption Judge Mayer is concerned about.
A thorough review of Judge Mayer’s analysis, combined with the fact that it was a concurring opinion, shows that the software patent is not dead. The current Alice framework directly addresses most of Judge Mayer’s concerns. Looking at the underlying reasons for Judge Mayer’s arguments suggests he is simply frustrated with the large number of bad software patents he sees on a regular basis.
There Is No Cookie-Cutter Solution
We should not spend much time waiting for a clearer standard on patenting software from the Supreme Court. Many recent cases seeking such guidance have been denied certiorari.[7] This is likely because the patent system has already learned first-hand the consequences of bright line rules. In its 2008 search to find a predictable test, the Federal Circuit declared the Machine or Transformation test as the standard for patent eligibility under §101.[8] While the Machine or Transformation test seemed to be in line with Supreme Court precedent, it had tremendous unintended consequences. The Machine or Transformation test led to numerous patents awarded merely because a conventional abstract idea was performed on conventional computer hardware. Today, many similar patents are regularly invalidated because implementing an abstract idea on a generic computer is not patent eligible. While many suggest that the sheer number of patents being invalidated is a sign of bad things to come, or worse, that software and its effects are not patent eligible, the fact that these patents are being invalidated is actually a good sign. The heightened number of invalidated patents is an indication that a lot of ineligible patents were issued under a system that hand-cuffed both patent examiners and the courts. The patent system is purging itself of patents that slipped through the system under the Machine or Transformation test.
The Supreme Court has long “warn[ed] …. against” interpreting Section 101 “in ways that make patent eligibility depend simply on the draftsman’s art.”[9] Trying to give a definition to the term “abstract idea” or a clear test on patent eligibility under §101 would do just that. Given the Alice framework, it’s clear that software patents will continue to be granted based on how well a patent prosecutor can define the invention so that it is not simply an “abstract idea.” A clear test with bright line rules and definitions would handcuff patent examiners and the courts for years, and once again set back the patent system.
For the purposes of §101, the want of predictability is outweighed by the need of flexibility. Patent law exists to promote the progress of science and useful arts. Scientific progress is unpredictable. An overly rigid legal system will only “impede innovation more than it would tend to promote it.”[10] Moreover, “Section 101’s vital role…is to insure that patent protection promotes, rather than impedes, scientific progress and technological innovation.”[11] The current application of the patent eligibility standard is working; no clear test is needed.
Summary
In trying to address new technology, the Federal Circuit used an inflexible rule to interpret Section 101. Since then, the Supreme Court has made determinations under Section 101 more flexible, which has led to large-scale purging of many patents that should never have been issued. The Supreme Court would not have gone through Bilski, Mayo and Alice, if software were ineligible for patent protection. Instead, the Supreme Court appears to be trying to mold a flexible set of rules that can keep pace with innovation. Another inflexible rule would simply set the patent system back again. The software patent is alive and well. It is merely being held to the same standard as all other areas of technology.
[1] Two Years After Alice: A Survey of The Impact of a ‘Minor Case’ (Part 1), Bilski Blog, June 16, 2016, available at: http://www.bilskiblog.com/blog/2016/06/two-years-after-alice-a-survey-of-the-impact-of-a-minor-case.html.
[2] Intellectual Ventures I LLC, v. Symantec Corp, 2015-1769, at 24-25 (Fed. Cir. 2016).
[3] Two Years After Alice: A Survey Of The Impact Of A ‘Minor Case’ (Part 1), Bilski Blog, June 16, 2016, available at: http://www.bilskiblog.com/blog/2016/06/two-years-after-alice-a-survey-of-the-impact-of-a-minor-case.html.
[4] Intellectual Ventures I LLC, v. Symantec Corp, 2015-1770 at 3 (Fed. Cir. 2016) (Mayer, C. J., concurring).
[5] Id. at 10.
[6] Id. at 12.
[7] Ultramercial, LLC et al. v. Wild Tangent, Inc. 772 F. 3d 709 (Fed. Cir. 2014) (cert. denied).
[8] In re Bilski, No. 2007-1130 (Fed. Cir. Oct. 30, 2008).
[9] Alice Corp. v. CLS Bank Int’l, 134 S. Ct. 2347 (2014).
[10] Mayo Collaborative Servs. v. Prometheus Labs., Inc., 132 S. Ct. 1289 (2012).
[11] I/P Engine, Inc. v. AOL Inc., 2013-1307, at*9 (Mayer, C. J., concurring) (“A robust application of section 101 ensures that the nation’s patent laws remain tethered to their constitutional moorings.”)