In recent court cases, judges have expressed concerns that claims for software-based inventions might attempt to preempt all possible uses of an abstract idea (see, for example Cybersource Corp. v. Retail Decisions, Inc., 654 F.3d 1366 (2011), CLS Bank International v. Alice Corp. (Fed. Cir. 2013) (en banc), ULTRAMERCIAL v. HULU , LLC (Fed. Cir. 2013), all of which we have blogged about recently). Generally, these inquiries have been directed to 35 USC §101, which governs patent eligibility of subject matter. Even when software-based claims have recited tangible subject matter, such as a system (e.g., a computerized system or a specially programmed computer) or computer-readable media (e.g. a CD or a memory with instructions), or a method that executes on a computer, rulings have come down that sometimes these sorts of claims are ineligible subject matter for a patent, and sometimes these sorts of claims are patent eligible. 35 USC §101, as discussed previously, forbids patenting of an abstract idea. Equations, and algorithms in isolation, are often considered abstract ideas. In rulings where a claim reciting tangible subject matter is found ineligible under this statute, 35 USC §101 is interpreted as extending to forbidding some claims that are based on an abstract idea. The question then becomes, how close to an abstract idea could a claim be, or how far away from an abstract idea should a claim be, in order for the claim to be found eligible under 35 USC §101. Relative uncertainty in this area is unsettling for the patent community, I believe.
One possible approach as to whether a claim is attempting to preempt all possible uses of an abstract idea is to utilize 35 USC §112, which governs definiteness of a claim, and enablement in the specification. A claim that truly is attempting to preempt all possible uses of an abstract idea should be rejected or overturned under 35 USC §112, because no specification could possibly enable “all possible uses of an abstract idea”. No human being, no known book, library or database could ever disclose all possible uses of an abstract idea, present or future, because the future is not yet known. And, inventors are so clever that, surely, it is possible some inventor will come up with a new use of whatever abstract idea is disclosed in a present patent application.
On the other hand, a claim that is properly enabled in the specification and that satisfies 35 USC §112 as to definiteness as well, and that recites tangible subject matter rather than solely an abstract idea, should go on to be tested against 35 USC §102 and 103 as to novelty and obviousness. In this view, 35 USC §101 becomes a basic test. Is the claim claiming an abstract idea itself? Or, is the claim claiming a concrete reduction to actual practice, based upon an abstract idea? Viewed this way, 35 USC §101 is very clear as to patent-eligible subject matter.
Testing a claim against 35 USC §101 then becomes straightforward. If the claim is for an algorithm itself, or an equation, or an idea, in isolation, the claim is not patent-eligible subject matter under 35 USC §101. If the claim is for a tangible article, including a specially programmed computer, a tangible media, a machine, or a method that operates on a specially programmed computer, the claim is patent-eligible subject matter under 35 USC §101, even though it be based in some manner on an abstract idea. We then go on to test whether the claim is overbroad, and unsupported in the specification, such as when a claim attempts to preempt all possible uses of an abstract idea. When the latter is the case, the claim is rejected or found invalid under 35 USC §112. I believe this proposed set of interpretations and sequences of inquiries would satisfy the justices’ concerns seen in the above-referenced cases, and elsewhere, and would clarify the situation regarding patentability of software-based inventions. In a subsequent blog, we will test this proposal against some blackbox machines, and see how rulings might go. Meanwhile, any comments?