“There should be no serious question that computer-implemented inventions such as software constitute patent-eligible subject matter under § 101,” Paul Clement wrote in a brief filed on behalf of IBM to the Supreme Court in 2014. Ultimately, the IBM brief would argue that the abstract idea doctrine is unworkable, which it is. Sadly, nearly 18 months after the Supreme Court’s landmark decision in Alice v. CLS Bank we are no closer to having a working understanding about when and under what circumstances software is patent eligible.
One might think that just about everything that can be said about Alice has already been said, but that is unfortunately not the case. Patent attorneys and innovators are struggling to understand what needs to be done to obtain and defend patent claims. Complicating this matter is the unfortunate truth that there is simply no way to reconcile all of the Supreme Court’s patent eligibility cases into a cohesive test that can reliably and predictably deal with the wide array of innovations that incorporate software.
We know that the Supreme Court did not say that software is patent ineligible, and we know that software patents are referred to specifically in the Patent Act in a variety of locations. In fact, when Congress passed the America Invents Act (AIA) a prohibition on the patenting of tax strategies was included. Also included, however, was a specific caveat that said that the tax strategy exclusion did not pertain to computer programs employing a tax strategy, which seems rather conclusive proof that Congress understood software to be patent eligible and wanted to specifically keep software, at least tax related software, patentable.
Unfortunately, real mischief has taken place since the Alice decision by the Supreme Court. The Federal Circuit, numerous district courts, the Patent Trial and Appeal Board (PTAB) and patent examiners at the United States Patent and Trademark Office (USPTO) have used Alice to reject software patent claims as being patent ineligible. Indeed, the Alice decision has infused a great deal of uncertainty into the law of patent eligibility. The level of uncertainty can perhaps best be exemplified by the fact that the Federal Circuit issued what appear to be diametrically opposed opinions in Ultramercial and DDR Holdings. See Federal Circuit finds software patent eligible and Arbitrary and Capricious: Judge Lourie’s Flip-Flop in Ultramercial.
Numerous patents have been lost with claims invalidated as being patent ineligible in the wake of Alice. This has negatively affected patent valuation, rendering many patents worth far less if not completely worthless. This has led some commentators to lament the toxicity of the patent asset. Companies invest in R&D with the expectation of future revenue and the entire life cycle of innovation is at risk.
As odd as all the machinations over software patents have been, perhaps the most bizarre is how the courts continue to try and make a distinction between a general-purpose computer and a specific purpose machine. To those familiar with software a distinction between a general purpose computer and a specific purpose machine is at best a distinction without a difference. The real value in software is that it operates across platforms, yet interoperability and compatibility seems to make the resulting software less likely to be patent eligible because it works on any type of machine (i.e., a general purpose computer versus some specific purpose machine). Such a distinction strikes me as ridiculous, and yet sadly descriptive of the peculiar state of the law relative to software patent eligibility.
The industry is nervous, and increasingly discussing the possibility of legislative reforms to address patent eligibility, which many believe is necessary. Numerous industry groups have formed patent eligibility working groups with the intent on drafting possible statutory language for a future reform of Section 101, but everyone is being cautious. If an attempt to reform patent eligibility legislatively were to go awry that could easily put the industry in an even more precarious position. These efforts will likely remain in committee within industry organizations and in private discussions pending what the Supreme Court may do next. Most expect the Court to take another software case at some point, perhaps sooner rather than later. There is also some hope that Ariosa v. Sequenom will potentially be a vehicle for greater certainty. See Petitioning the Federal Circuit to Reverse Court on Patent Eligibility.
The fundamental problem created by Supreme Court software related patent jurisprudence is that on some very basic level everything can be characterized as an idea. This truism, however, does not mean that software is itself an abstract idea. Software instructs a machine to operate in a specific way to accomplish a specific task. Implementing workable software is anything but trivial, but that nuance is often lost on those without technical training who seem to think virtually anyone can create software.
I personally believe it is only a matter of time before we have greater clarity on software patent eligibility, and that clarity will result in software being more readily patent eligible. In the meantime, however, clients need to make decisions. What we do know is that the law of software patentability has changed over the last 30 years about every two to three years. Given that a patent application filed today will almost certainly not receive a First Office Action on the Merits for two to three years that means the law will evolve at least once, perhaps twice, before new patent applications filed today are examined.
Of course, filing patent applications must be done with an eye toward the future. At the moment most patent attorneys are recommending longer more detailed disclosures that painstakingly describe the technology in ways that makes the overall system seem more tangible, and thus likely more familiar to judges.