After a decade of uncertainty about business-method patents, the U.S. Court of Appeals for the Federal Circuit has weighed in once more, leaving us, finally, in a definite state of...uncertainty. This past October, in In re Bilski , the court ruled that a process or method can be patented if it is tied to a particular machine or if it transforms a particular article into a different state or thing. Legal scholars and patent attorneys are already furiously disagreeing about what that means. Only two things are certain: the door is still ajar for business-method patents, but it’s not as open as it used to be.
The case most often cited for opening the door to business-method patents is State StreetBank & Trust v. Signature Financial Group Inc. , decided in 1998 by the same court. That case, however, involved a fairly complex computer program, and all the court really said was that the validity of the patent should not turn on whether the ”subject matter does ’business’ instead of something else.” Even the poster child of business-method patents, the Amazon ”one-click” patent, which was in the patent pipeline well before State Street , involved software operating on a specially designed client-server system.
So even in the years preceding State Street , while many patent attorneys, including myself, regularly procured patents covering software and the Internet, what we actually pondered was the real ”business method” question: could you patent something that had no technology component at all? Suppose, for example, a client ”invented” a way to pick winning stocks—if a stock has attributes A, B, and C, then you should buy it. Could a patent be used to protect that method? State Street said nothing about such a patent, and neither had any other court case.
In 1997, two energy-industry executives, Bernard Bilski and Rand Warsaw, filed a patent application for a method of managing, in the face of rising and falling demand, the cost of a commodity like coal. The method doesn’t need computers—in fact, they aren’t even mentioned in the patent application. Essentially, the patent would protect the idea of an intermediary who buys coal from mining companies at a fixed price, insulating coal firms from falling prices. The same intermediary would sell coal to power plants at a second fixed price, thus insulating the power plants from price increases. The intermediary has thus hedged its risk; if demand and prices skyrocket, it has sold coal at a disadvantageous price but bought coal at an advantageous price, and vice versa if prices fall.
The U.S. Patent and Trademark Office has rejected the application as a nonpatentable business method. Reading its rejection, however, would lead no one to a particularly clear understanding of what is and what is not patentable.
The problem lies in the patent law itself, which defines as patentable ”machines,” ”compositions of matter,” and ”processes.” Machines and compositions of matter are usually easy to identify. But what is a process? A new metallurgy technique? A specific treatment of silicon wafers? Definitely. But what about my client’s process for picking stocks? The fact that it is performed manually is of no import—you can patent a way to mix two compounds together to produce a new chemical substance even if it’s by hand. To distinguish between nonpatentable processes and patentable processes, you have to interpret the law as possessing a requirement that is not explicitly stated, namely that patents are for technological or scientific methods only—processes that transform something physical. This, in effect, is what the Patent Office did.