In addressing claims directed to medical diagnoses, the Supreme Court's opinion in Mayo Collaborative Services v. Prometheus Laboratories, Inc. may actually lend some clarity to questions of patentability pertaining to software patents. True, a clearer understanding of software patentability might not be the most obvious take-away from Prometheus. And we are still a long way from reliable predictability for litigants concerning questions under 35 U.S.C. § 101. But Prometheus did feature some relatively straightforward analysis, and certain clear statements of law, concerning Section 101.
I will not repeat the able summaries of the Court's opinion that others have already provided. But I would like to share ideas about some general lessons to be drawn, and extrapolated to the realm of software patents. You may not agree with all of my thoughts, but these are certainly the things that any practitioner should be thinking about.
The Court rejected the idea that "virtually any step beyond a law of nature itself should transform an unpatentable law of nature into a potentially patentable law of nature" under Section 101. Invalidity analysis based on prior art, or alleged indefiniteness cannot, the Court said, replace Section 101. This is an important statement because many people, such as the majority of the Federal Circuit panel in MySpace, Inc. v. Graphon Corp., have argued that Section 101 should not be a threshold test, but rather should be a test of last resort when all other invalidity analyses have been exhausted. While the Court did not explicitly address the "threshold test" question, the discussion certainly reinforces that Section 101 should not be relegated to second-tier status in invalidity proceedings (although some may see irony in the Court's statement that "shift[ing] the patentability inquiry to [Sections 102, 103, and 112] risks creating significantly greater legal uncertainty").
Justice Breyer's opinion strikingly went through Prometheus' representative claim element by element to look for patent-eligible subject matter. Or perhaps it is striking that some lower court opinions have not done this, a recent example being the Federal Circuit's opinion in Ultramercial v. Hulu. I believe that the Supreme Court's element-by-element approach has a lesson for software patent drafters. While the court repeated the well-known proposition that "post-solution activity" (such as providing output to a display) does not patent eligibility make, the implication of the court's approach is that a single claim element can pass the claim over the Section 101 threshold. If you can put your software into a system, for example, and if that system includes includes an element that is patent-eligible, then perhaps your claim will likewise be patent-eligible -- regardless of what other steps it recites. However, as discussed below, that claim element must also be non-obvious and novel.
Concerning the steps in Prometheus' claim, the Court stated that "to consider the three steps as an ordered combination adds nothing to the laws of nature that is not already present when the steps are considered separately." Then, discussing the venerable case of Diamond v. Diehr (holding patentable a method for curing and molding rubber that included an algorithm), the Court stated that "[i]t was nowhere suggested that all [of the claimed] steps, or at least the combination of those steps, were in context obvious, already in use, or conventional." I won't go on, but there are similar statements in the Court's Bilski opinions, and in various Federal Circuit opinions under Section 101. It may be that it is enough to simply recite any old physical element in a set of software steps -- and sometimes that's all you can do -- but claim language (at least in dependent claims) drawn to physical and novel, non-obvious elements is plainly highly advisable.
Courts commonly point to the breadth of claims as evidence of their invalidity under Section 101. The Prometheus Court, like various Federal Circuit opinions, and like the two main opinions in Bilski, emphasized claim breadth. The more detail that is included in a claim, the more difficult it will be for the Court to say that the claim is drawn to an abstract idea, particularly under the element-by-element approach of Prometheus. The perhaps prohibitive downside to this recommendation is that it is probably an all-or-none strategy; if you include several broad independent claims along with a more narrow "picture" claim, the picture claim will probably get swept into the analysis with the broad independent claims.
Bilski clearly stated that satisfying the "machine-or-transformation" test does not necessarily render a claim patent-eligible. Prometheus does nothing but reinforce that proposition. However, Prometheus also reinforces that the machine-or-transformation test always has effect the other way: if a court can say that a claim does not satisfy the test, the court will hold the claim not patent-eligible. This probably goes without saying ten times over, but, in a software case, never file a claim without a machine (unless you have something like the rubber-curing process of Diamond v. Diehr, though such a claim would quite probably require a machine in any event).
Prometheus was about laws of nature, while Bilski was about abstract ideas. Think about the "law of nature" claimed in Prometheus (a correlation of chemical measurements in blood with recommended drug dosages) compared with the "abstract idea" claimed in Bilski (a way to hedge risk in commodities transactions). Prometheus' correlation is a naturally occurring relationship discovered and used by the inventors. Bilski's method of hedging risk is not susceptible to such a neat definition. Some people in the medical field are happy with the result in Prometheus, and some are not, but at least they have some degree of certainty. Software claimants have no similar benchmarks.
It was unanimous (again).
For the second consecutive time, all nine Supreme Court justice (and a different nine in each case, to boot), have found claims not patent-eligible. It will be interesting to see whether the Court will grant certoriari in the Ultramercial, LLC v. Hulu, LLC case.