Last October 30 was Mischief Night. To those not familiar with the occasion, Mischief Night is the night before Halloween when young hooligans gallivant around your community perpetrating various random acts of petty violence – toilet-papering trees, egging cars and other such annoyances. Last year, on Mischief Night, the Court of Appeals for the Federal Circuit (CAFC) got into the act, and dropped a bombshell on the software and business method patentee communities with their announcement of Bilski.
In a nutshell, Bilski shrinks the scope of patentable subject matter and excludes most useful software and business method claims from patent protection. The instrumentation chosen by the court to strike this dizzying blow was the broad adoption of the Supreme Court’s machine-or-transformation test. That test, announced in Gottschalk v. Benson and later repeated in Diamond v. Diehr, states that the best way to determine whether a patent applicant has made an impermissibly broad process claim is to ask whether the claimed process is tied to a particular machine, or whether it transforms an article into a different state or thing. If it does, it is sufficiently narrow and therefore patentable. If it does not, then the claim is invalid as claiming an abstract idea/fundamental principle. Bilski makes two important additions to this test: (1) it states that the test is not the best way, but rather the only way for a process to be patentable, and (2) it makes every effort to foreclose clever draftsmen from squeezing impermissible claims into the machine-or-transformation rubric – stating that recitations of machines or transformations that do not go the heart of the invention or appear to be superfluous “extra-solution activity” will not be enough to transform an impermissibly broad claim into a patentable one. In a tangential move, the Bilski court also holds that field-of-use restrictions are insufficient to transform unpatentable claims into permissible ones.
The reason this holding has struck software and business method patentees with such force is because it is a departure from over a decade of precedent and because it endangers billions of dollars’ worth of issued patents and associated value. The CAFC’s State Street case has been credited with announcing that business methods were patentable (including software). Since that case was decided in 1998, several tests had been developed by the CAFC to try to further capture precisely what it is that makes a process patentable or not. Bilski abandons each of these additional iterations. The “Useful, Concrete Tangible Result” test – gone. The “Physical Steps” test – irrelevant. The Freeman-Walter-Abele test? Also gone. Post-Bilski, only the machine-or-transformation test will do. This causes great consternation for software developers like Microsoft because it is likely, though not a surety, that a general-purpose computer – like your computer desktop – is not a “particular machine” within the meaning of the test. Ergo, much of Microsoft’s patent portfolio relating to general-purpose computers is now probably unenforceable. Financial services companies are equally concerned because many of the innovations they have patented over the past 10 years have dealt with the manipulation of intangible constructs on a general-purpose computer. In sum, Bilski is nearly fatal to business method patents and severely limits a software developer’s ability to protect their programs via patent.
Despite the fearful reaction of many within these communities, there are some who believe that Bilski is not as menacing as it appears. These Bilski detractors point out that the CAFC did not categorically outlaw software or business method patents. They point to Bilski’s footnote 23, which so states and opine that patent drafters will find ways to draft around the “extra-solution activity” language soon enough. These advocates also argue that Bilski’s main premise – that Diehr and Benson command that the machine-or-transformation test be the test – is flawed, and therefore will be overturned by the Supreme Court. More specifically, they note that Benson, which announced the test, specifically stated that the Court was not holding “that no process patent could ever qualify if it did not meet the requirements of [the test]”. They argue that this hedging language “commands” that the test be treated as a judicial safe harbor rather than an exclusive gateway to patentability. They acknowledge that this language is missing from the subsequent Diehr decision, a fact which Bilski hangs its proverbial hat upon, but argue that this omission is insufficient to justify the sweeping precedential changes that Bilski makes.
These arguments have merit and Bilski may very well be overturned by the Supreme Court. Until that happens, however, software and business method patentees should be cautious and try to wait out any pressing infringement situations. In sum, though not perpetrated by intellectual hooligans, Bilski will go down as a trick exacted against software and business method patentees if left unchanged. Hope abounds, however, that cert and clarity are forthcoming.
Thomas F. Gillespie III is an associate on the in-house counsel team at the J. Craig Venter Institute in Rockville, Maryland. He concentrates his practice in patent law.