Yesterday Bloomberg News published an article describing new contract review software created by JP Morgan (NYSE: JPM) that is capable of doing in seconds the same work that it would take a large team of lawyers 360,000 hours to complete. The author — Hugh Son — explains:
The program, called COIN, for Contract Intelligence, does the mind-numbing job of interpreting commercial-loan agreements that, until the project went online in June, consumed 360,000 hours of work each year by lawyers and loan officers. The software reviews documents in seconds, is less error-prone and never asks for vacation.
Clearly, this extraordinary software solution for engaging in tedious contract review is nothing more than an abstract idea and is not the type of thing that can be patented in America. The United States Supreme Court has put an end to revolutionary innovations such as COIN from being patented, and if the hard working patent examiners at the United States Patent and Trademark Office make a mistake and issue a patent on such a ridiculously simple innovation that a second year engineering student could clearly have programmed over a weekend while sipping latte’s at the corner coffee shop the PTAB, some district court or the Supreme Court will step in and set the record straight. Even certain panels of the Federal Circuit would undoubtedly find COIN to be patent ineligible because, after all, the software is just doing what humans could do anyway.
Yes, the old, extraordinarily ridiculous argument that humans could perform the task of the software so there isn’t anything special to see here — move along. That was fundamentally why the Federal Circuit found the claims in CyberSource Corp. v. Retail Decisions, Inc. to be patent ineligible. Judge Dyk, writing for a unanimous panel that included Judges Bryson and then Judge (now Chief Judge) Prost, wrote: “It is clear that unpatentable mental processes are the subject matter of claim 3. All of claim 3’s method steps can be performed in the human mind, or by a human using a pen and paper.”
It is hard to see how software that performs the mental review function of a lawyer, just much faster, could be patentable in view of CyberSource. Of course, that doesn’t say or suggest anything about the remarkable innovation that JP Morgan has achieved. It does, however, speak volumes about the idiotic tests and application of those tests by judges who think they have all the answers and seem absolutely hell bent on treating all software as if it is ubiquitous. A simple app on your phone to track your caloric intake is just the same as IBM’s Watson that won on Jeopardy, or IBM’s Deep Blue that beat Chess Grand Master and World Champion Gary Kasparov, or JP Morgan’s COIN. After all, a second year engineering student can simply code each over a weekend in a coffee shop while sipping on a latte.
For those unfamiliar with the repeated engineering student reference, Justice Kennedy of the United States Supreme Court actually said in oral arguments during Alice v. CLS Bank that the software could be written by a second year engineering student over a weekend. In what can only be characterized as one of the worst ever responses by an advocate during an oral argument at the Supreme Court the attorney representing the patent owner agreed. Unbelievable! The answer is NO! No Your Honor, it cannot be programmed by a second year engineering student over a weekend, and that question shows a complete and total lack of familiarity with the scope of even a simple software coding project. It would be difficult, if not impossible, for anyone to code that simple app for your smartphone to track your caloric intake in a weekend, at least if you wanted it work without bugs and so that it was secure and not an identity theft open door. And second year engineering students have learned little or no computer programming anyway. So the question and the answer clearly demonstrate to everyone who knows anything about technology, software and/or engineering that those who are making these decisions think they know it all but know absolutely nothing.
So how would an Alice/Mayo framework analysis for COIN proceed? The first question (commonly referred to as Step 1) is whether the patent claim covers an invention from one of the four enumerated categories of invention defined in 35 U.S.C. §101 (i.e., is the invention a process, machine, article of manufacture, or composition of matter). It would seem that even the most recalcitrant decision makers no longer question Step 1, if for no other reason because it is unnecessary to fudge on the obvious when the next two questions provide all the subjective leeway they need to kill any patent claim they want.
The second question (commonly referred to as Step 2A), where the Alice/Mayo framework truly begins, requires the decision maker to ask if despite the claim being deemed patent eligible under the law, whether the patent claim seeks to cover one of the three specifically identified judicial exceptions to patent eligibility. At the moment there are only three identified judicial exceptions, which are: laws of nature, physical phenomena and abstract ideas. With software the judicial exception at play is always the abstract idea exception, which the Supreme Court and Federal Circuit have steadfastly (and rather proudly) refused to define. If the claim does NOT implicate one of those judicial exception then the claim is patent eligible. This is, of course, subjective. In this situation pretty much any of the Administrative Patent Judges of the PTAB, most district court judges, most patent examiners, and a handful of Federal Circuit Judges (i.e., Judges Dyk and Mayer for certain) would find that the abstract idea is nothing more than reviewing contracts, which is something that has been done by human beings for hundreds of years without the assistance of computers.
The abstract idea is identified by those who will find the patent claims patent ineligible without interpreting the claims, and many times without reading the claims. Of course, black-letter law requires full consideration of the claim, the specification and prosecution history in order to determine what a claim covers, but for some reason when reaching a patent eligibility decision very little more than the preamble to the claim is relevant. There is almost never a review of the specification to even see if any terms or concepts have been defined, as is absolutely required to interpret a claim. And there is never any review of the prosecution history. So the abstract idea is identified in some ether or vacuum based on the most absurdly broad articulation of what the decision maker wants the reader to believe the patent applicant or owner is seeking or obtained, as the case may be.
In the case where the patent claim seeks to cover a judicial exception to patent eligibility, the final question (commonly referred to as Step 2B) asks whether the inventive concept covered in the claimed invention added “significantly more” than the judicial exception, or whether the claimed invention did not add “significantly more” and, therefore, was seeking to merely cover the judicial exception. Again, the Supreme Court and the Federal Circuit have never defined what “significantly more” means. This frequently gets characterized as the hunt for the inventive concept, and if you think the abstract idea prong is subjective wait until you get more acquainted with this prong of the Alice/Mayo framework. It is really impossible to breakdown this prong with any objectivity because there is never any in the decisions. It is as subjective a criteria as you will ever see in the law, akin to the distinction between pornography and obscenity — you’ll know it when you see it.
Of course, that standard is not helpful when it comes to indecency, and it is absolutely useless when it comes to matters of business law that demand certainty. And let’s make no mistake about it, software patents are about business, sought by businesses and the lack of certainty is harming those businesses, the U.S. economy and driving companies overseas for protection.
To all those who will comment that there is no way to know whether claims to the JP Morgan COIN software would be viewed as patent ineligible without seeing what, if anything, might be claimed, my question is this: Do the claims even matter? As far as I can tell, the claims are completely incidental to a patent ineligibility analysis. All that matters is the vague, overbroad description of the abstract idea concocted by the decision maker. After all, how else could an MRI machine be considered to be an abstract idea? You’d think that even PTAB judges would realize that in a dark room without any lights they would trip and fall over that allegedly abstract MRI machine, which would make it not so abstract after all, right?
In any event, congratulations to JP Morgan on creating something that seems quite revolutionary. Too bad you won’t get any useful patent protection in the United States. I recommend you consider China, Europe and other patent friendly jurisdictions around the world. The U.S. seems closed for business with respect to these types of paradigm shifting innovations.