Should you patent or not?
When it comes to software, it's as clear as mud
One of the most common questions I get when meeting with new clients is whether their invention is patentable.
To be patentable, an invention must satisfy three basic criteria: it must be the proper subject matter; it must be new; and it must not be obvious. For most of the last 20 years or so, the first criteria was an easy call.
If the invention was a useful process, a machine, an article of manufacture or a composition, it was patentable subject matter. If the invention as described and claimed in a patent application was also found to be new (novel) and not an obvious derivation from that which was already known, then a patent would issue.
Unfortunately, when lawyers and the courts get involved, even the straight forward plain language of a statute can be made to appear confusing. Those of us over 40-years-old might recall it when President Clinton, a soon-to-be disbarred lawyer, argued that the word “is” had potential several different meanings in relation to statements he made concerning the Monica Lewinsky scandal.
If a lawyer can, with a straight face, argue about what the meaning of the word “is” is, imagine what a skillful lawyer could do to something like the patent statute. The problem is, when lawyers and the courts play word games, they soon find they have worked the law into an impossibly-difficult legal jigsaw puzzle.
This is exactly the situation that developed in the patent world concerning software-related inventions. For most of the 70s and through the 80s, software methods were generally not considered patentable subject matter for a whole host of reasons: they were drawn to abstract ideas; they were the equivalent of mental steps; they pertained to business methods and did not have technical character; and they were mere algorithms.
A hole formed in the software patent dike in 1981 when the Supreme Court declared in a case concerning DNA-related inventions that “anything useful under the sun invented by man is patentable.”
It seemed a simple enough pronouncement, but the Patent Office and the lower courts continued to struggle with the notion for the next 17 years until a couple of cases in the federal court of appeals finally directly applied the DNA-related Supreme Court holding to software. The decisions coincided with the dot.com boom, releasing a biblical-size flood upon the patent office and later on the courts that had to interpret this new breed of patent.
At the time as a law school student, I authored a law review article concerning the march towards software patentability and concluded that the matter was finally and correctly settled: “[it] appears the … debate is finished in the United States. Software is to be treated no differently than any other invention.”
I spoke far too soon. Clever lawyers and sympathetic judges immediately began rebuilding the dam and just under two years ago the Supreme Court weighed in, dealing a mortal wound to a high percentage of the software-related patents issued over the last 15 years, and effectively preventing patent protection for thousands of new software innovations.
While patenting software innovations isn’t dead, the uncertainty of whether a particular software-related invention is patentable subject matter has become highly uncertain. Also, the average cost of obtaining a software-related patent for those applications that survive the subject matter inquisition is considerably higher than it was even five years ago.
The two-step test laid out by the Supreme Court to determine whether a software-related invention is patentable subject matter is rather simple on its face, but almost nonsensical in practice.
First, a patent examiner must determine whether the claims are directed to a patent-ineligible concept, and then if claims are directed to patent-ineligible concepts, he must determine whether the claim’s elements, considered both individually and as an ordered combination, transform the nature of the claims into a patent-eligible application.
The Supreme Court, however, did a poor job in defining what is a “patent-ineligible concept,” providing some broad categories but not defining them very well. The Patent Office moved in and issued guidelines to fill in the blanks.
Even the Patent Office became confused. In guidelines drawn to natural products, which have also come under scrutiny as not being patentable subject matter, the Patent Office provided an example of the proper application of the two-step test.
The Office hinted gunpowder would not be patentable subject matter as a natural product, stating, “gunpowder as recited in the claim [is] not markedly different from what exists in nature.” The document in question was issued by a deputy commissioner at the Patent Office and was presumably vetted before being issued.
The office ultimately recognized the error and issued a new example, this time finding that gunpowder would be patentable subject matter if it were newly invented. The mistake seems harmless enough, but imagine if you are one of hundreds or thousands of potential patentees that suffered a rejection of your patent application because the Patent Office was trying to sort out the law.
Where does all of this leave the inventor or company seeking to protect its software-related intellectual property? Unfortunately, this is all about as clear as mud.
I think the Patent Office overreacted to the Supreme Court holding, interpreting it much more narrowly than was intended and that over time more and more software-related patents will be found to be proper subject matter. In fact, some lower court decisions are just starting to trickle in seemingly rejecting the Patent Office's narrow interpretation.
My advice to inventors of software-related products and processes is to assess the potential value of the innovation and determine whether the high cost of patenting this type of invention is worth the expense.
For those that are, I often recommend filing a provisional patent application first to gain an extra year before the subsequent non-provisional patent application is examined by the Patent Office. In the two to four-year wait for the Patent Office to take up the application, hopefully the interpretation of the law will have loosened a bit and what might be an invention of dubious merit under the current subject matter paradigm will sail through without raising an eye. Time will tell.