By Michael Borella --
On December 6th, the Supreme Court granted certiorari to a case that presents the question of "[w]hether claims to computer-implemented inventions -- including claims to systems and machines, processes, and items of manufacture -- are directed to patent-eligible subject matter within the meaning of 35 U.S.C. § 101 as interpreted by this Court?" Thus, for the third time in five years, the highest court will once again review the patent-eligibility of so-called software and business method patents.
This case was likely destined the Supreme Court once the Federal Circuit released its six-opinion en banc decision this May, demonstrating a severe intra-circuit split. Judge Lourie's concurrence read Justice Breyer's 2012 Mayo v. Prometheus opinion in view of former Justice Stevens' 1978 Parker v. Flook decision, the latter a case that some thought had been in part overturned by 1980's Diamond v. Diehr and 2010's Bilski v. Kappos. In Mayo, however, Justice Breyer resurrected the "inventive concept" analysis of Flook, and its propensity to blur the lines between the inquiries of patent-eligible subject matter, anticipation, and obviousness.
In applying this analysis in CLS Bank, Judge Lourie set forth a procedure for analyzing claims that incorporate abstract ideas. Since then, it has been roundly criticized for failing to produce an objective, workable standard. Chief Judge Rader wrote separately, taking an approach that attempted to apply the inventive concept analysis while keeping § 101 separate from considerations of prior art. Other panel opinions took different methodologies, relying on procedural issues to tip the scales, or advocating a dramatic reduction in scope of the judicial exceptions to § 101.
Regardless, the en banc panel did little to clarify the bounds of patent-eligible subject matter, and may have muddied the waters even further. Therefore, it is not at all surprising that the Supreme Court has taken up this case. Hopefully, the Justices will clarify whether prior art should be considered in a 35 U.S.C. § 101 analysis, whether such an analysis should be applied to claims as a whole or to claim elements on a piecemeal basis, and whether recitation of general-purpose computer hardware (either as a claim element or as the statutory type of the claim) makes a claim any less abstract. Or, being the Supreme Court, it is within their purview to decide the case on much narrower grounds.
Now that the case has been fully briefed, we can look to the parties for the positions that they will likely be taking during oral argument.
Brief Overview of the Case
In 2007, CLS Bank filed a declaratory judgment action against Alice, contending that, among other things, Alice's U.S. Patents Nos. 5,970,479, 6,912,510, and 7,149,720 were invalid under § 101. Alice filed counterclaims alleging infringement of these three patents. Later, Alice's U.S. Patent No. 7,725,375 was added to the action, and the parties respectively asserted invalidity and infringement contentions for this patent as well.
All four patents are from the same family and "share substantially the same specification." The Federal Circuit plurality described the claimed subject matter as:
[A] computerized trading platform used for conducting financial transactions in which a third party settles obligations between a first and a second party so as to eliminate "counterparty" or "settlement" risk . . . . Settlement risk refers to the risk to each party in an exchange that only one of the two parties will actually pay its obligation, leaving the paying party without its principal or the benefit of the counterparty's performance. Alice's patents address that risk by relying on a trusted third party to ensure the exchange of either both parties' obligations or neither obligation.
Between the four patents, method, computer-readable medium (CRM), and system claims were asserted. The parties stipulated that all claims, including the method claims, require "a computer including at least a processor and memory" and electronic implementation.
The District Court held that all asserted claims failed to meet the requirements of § 101, and thus were invalid. Particularly, the District Court concluded that the method claims were directed to an unpatentable abstract idea, and that the CRM and system claims would similarly preempt all practical applications of this idea, despite those claims falling under a different statutory category.
On appeal, a Federal Circuit panel reversed the District Court on all counts, holding that all claims, including the method claims, were patent eligible under § 101. CLS Bank petitioned the full Federal Circuit for review, which was granted. Sitting en banc, seven of the ten judges overruled the panel and affirmed the District Court's ruling regarding the method and CRM claims. However, this majority did not agree on the rationale for their conclusion. Further, five of the judges found that the system claims were unpatentable, while the other five concluded that these claims passed muster under § 101. Additionally, eight of the judges indicated that the method, CRM, and system claims must rise or fall together.
Alice's Brief
After the en banc Federal Circuit decision, Alice found itself in a difficult position. To obtain a favorable Supreme Court review, it has to thread the eye of a needle by arguing that the Federal Circuit has left the law in a state of disarray, without suggesting that this disarray stems from the Supreme Court's own 2012 Mayo ruling. Also, Alice has to distinguish its method claims over those that the Court struck down in Bilski, despite a high-level similarity thereto.
Alice began its substantive analysis by noting that while the requirements of §§ 102, 103, and 112 must be met for a patent to be valid, subject matter eligibility was the solely the province of § 101. Alice quoted Judge Rich, "[a]chieving the ultimate goal of a patent [requires] having the separate keys to open in succession the three doors of sections 101, 102, and 103." Thus, Alice signaled early that it was going to lean on Diehr as the controlling precedent.
In presenting its claims, Alice wisely chose a system claim, claim 26 of the '375 patent, to discuss first. This claim explicitly recites a computer, coupled to a communications controller and data storage, which communicates with a separate device. Clearly, this claim recites tangible, physical hardware. Alice also presented claim 33 of the '479 patent as a representative method claim. However, this method claim is disembodied, and does not explicitly recite any physical hardware. Instead, Alice contended that "a person of skill in the art would understand the methods, as claimed, to require use of a computer." As noted above, this was also stipulated to by both parties in the district court proceedings. Alice also briefly discussed its CRM claims, noting that the media is what allows the claimed systems to use the claimed methods.
In its review of the case so far, Alice took the opportunity to poke at Judge Lourie's plurality opinion. For instance, Alice, pointed out that Judge Lourie "look[ed] past the claim language" when conducting his piecemeal inventive concept analysis, and re-emphasized that "claims must be considered as a whole." This sets the stage for Alice's later attempts to drive a wedge between Justice Breyer's Mayo analysis and the approach of the Federal Circuit's plurality.
Alice's substantive argument consists of three main parts: (i) defining the "abstract idea" exception to § 101 and contrasting it to patent-eligible disembodied concepts, (ii) reiterating that claims must be read as a whole, and (iii) asserting that, regardless of whether Alice's claims encompass abstract ideas, they are patent-eligible.
Alice began its discussion of abstract ideas by noting that, unlike laws of nature and physical phenomena, the Court has never clearly defined what it means by that term. However, in Alice's view, the Court has treated abstract ideas substantively the same as it has treated laws of nature. Both are fundamental truths that reflect "relationship[s] that [have] always existed" and therefore are "part of the storehouse of knowledge of all men . . . free to all men and reserved exclusively to none." In other words, abstract ideas were not created or invented by humans, and are instead fundamental aspects of the universe that belong in the public domain.
Discovery of an abstract idea, such as "mathematical formulas and the like," in and of itself does not warrant a patent for the discoverer. Instead, "[w]hen a patent claim recites such an abstract idea, it can be patented only if the claim includes other elements or steps that go beyond the fundamental truth itself, thus limiting the claimed invention to a specific application and preserving public access to the fundamental truth." Claiming a use of an abstract idea in a particular field does not lift the claim to the level of being patent-eligible, as no one is entitled to foreclose others' use of this public domain knowledge.
Alice painted a picture in which Gottschalk v. Benson and Flook were decided in order to avoid the preemption of an abstract idea. In these cases, the claimed abstract ideas were both algorithms, rather and mathematical relationships, but still were considered laws of nature. Moving on to Bilski, Alice couched that case's claims as also being an algorithmic abstract idea. Particularly, Alice pointed to the Court's statement that Bilski's "[c]laim 4 puts the concept articulated in claim 1 into a simple mathematical formula."
To contrast with abstract ideas, Alice introduced the notion of "disembodied concepts." In Alice's assessment, a disembodied concept does not fall into any of the four statutory categories of § 101, and instead is a "pure mental conception." All claimed inventions involve, in one way or another, disembodied concepts. The language of the claim is what makes these concepts statutory. Further, and important to Alice's argument, a disembodied concept need not be an abstract idea or a law of nature, and therefore it is inappropriate to conduct a preemption analysis on such a claim. Therefore, Judge Lourie's approach of asking whether such an "inventive [disembodied] concept" is wrapped up in a claim will always be answered in the affirmative, rendering his analysis to be excessively restrictive.
Alice moved on to a related topic, that all claims, whether reciting an abstract idea or not, must be read as a whole. Alice noted that the Court has found it "improper to look to the 'heart' or 'gist' of an invention, rather than the actual invention as described in the claim's language." Instead, all of the claim's limitations and terms must be considered, because this language defined the boundaries that the "PTO has actually found deserving of a patent and what the public may not use." Thus, as noted in Diehr, "[t]he 'novelty' of any element or steps in a process, or even of the process itself, is of no relevance in determining whether the subject matter of a claim falls within [§ 101]."
In order to drive this point home, Alice needed to synthesize the seeming discrepancies between Diehr and Mayo. In doing so, Alice noted that, like Flook, the claim under consideration in Mayo recited an algorithm for exploiting a law of nature, and that the claim covered all practical uses of this law of nature. Thus, the "inventive concept" analysis of Mayo does not advocate dividing the claim into abstract and non-abstract elements, because in Mayo the claim was considered as a whole. Thus, Alice purported that the Court in Mayo did not overrule Diehr, but instead "the 'inventive concept' inquiry asks whether additional steps in a claim that on its face recites a law of nature or fundamental truth do anything more than simply say 'apply it.'"
From all this, Alice concluded that computer-implemented inventions are necessarily physical embodiments, and that such an invention that would "preempt use of a mathematical formula or other fundamental truth -- as did the inventions in Benson and Flook -- is ineligible for patenting." However, computer-implemented inventions that do not exhibit such preemption are patent-eligible.
Alice set forth a two-part test for determining the patent-eligibility of computer-implemented inventions:
If the claim does not, on its face, recite a mathematical formula or other fundamental truth, it is eligible.
[I]f the claim recites a particular application of a mathematical formula, such that others would not be foreclosed from using the formula in other practical applications, it is eligible.
Alice reminded the Court that such a test is merely a threshold test for § 101 purposes. Sections 102, 103, and 112 prevent patents from claiming inventions too broadly or without enough specificity.
Finally, Alice argued that its claims were valid under § 101. Alice noted that unlike Benson, Flook, and Bilski, Alice's claims had been allowed and issued by the U.S. Patent and Trademark Office. Therefore, they enjoyed the presumption of validity, which saddled a high burden of proof on CLS Bank to establish otherwise.
Alice first argued that its claims are not directed to an abstract idea because they do not recite a mathematical formula, or a fundamental truth that can be reduced to a mathematical formula. Thus, by default, Alice's claims cannot preempt others' uses of abstract ideas. Alice contended that Judge Lourie held otherwise because Alice's method claims were drawn to a disembodied concept, and that he failed to appreciate the difference between such a concept and an abstract idea.
Alice also took issue with the lower court's "paraphrasing and dissecting the language of Alice's claims in its hunt for a disembodied concept, instead of reviewing the claims as a whole and as written." Alice blamed this approach for "untethering" the claims from the recited real-world application. Again, Alice emphasized that even their disembodied method claims were understood to require an unclaimed computer implementation, which Judge Lourie improperly ignored.
In the alternative, Alice asserted that even if its claims encompassed an abstract idea, they are directed to the specific application of using a computer as the intermediary for third-party escrow transactions. Alice pointed out that "a differently designed computer system could implement exchanges in other ways," and provided several examples thereof. Thus, at worst, Alice's claims are similar to those of Diehr, where an abstract idea is invoked, but the claim as a whole does not preempt all uses of that idea in any particular field.
Analysis
Alice's main difficulty is that its method claims are similar in both structure and content to those of Bilski, and the Court may not believe that it is appropriate to hold that Alice's claims meet the § 101 requirements when Bilski's did not. The Court might not defer to the stipulation that these disembodied method claims must be performed by a computer -- the claims, on their face, could be performed between humans, perhaps with the assistance of pencil and paper.
On the other hand, the Court should have a hard time concluding that Alice's system claims fail to pass muster under § 101, because doing so will put hundreds of thousands of similarly-claimed patented inventions at risk of being rendered invalid. It is unlikely that the Court will want to stir the pot to such a degree without a clear signal from Congress that software and business methods should be frowned upon.
Alice's abstract idea / disembodied concept dichotomy is a clever approach that could potentially be adopted by the Court in some form. It appeals to logic, and frames the Court's previous decisions in a consistent fashion. The wildcard is whether Alice's understanding of Mayo agrees with that of its author. Justice Breyer might respond to Alice's interpretation by saying, "That's not what I meant." Such a statement would throw the proverbial spanner in Alice's gears.
Lastly, students of computer science -- indeed anyone who has ever used a computer -- might be entertained by Alice's seven-paragraph argument that a computer is a physical machine. Alice has raised stating the obvious to a form of art. Too bad that such an exercise is necessary in 2014.
"Hopefully, the Justices will clarify whether prior art should be considered in a 35 U.S.C. § 101 analysis, whether such an analysis should be applied to claims as a whole or to claim elements on a piecemeal basis, and whether recitation of general-purpose computer hardware (either as a claim element or as the statutory type of the claim) makes a claim any less abstract."
Michael,
A very faint hope. Our Judicial Mount Olympus has created this muddle and as you correctly observe "this disarray stems from the Supreme Court's own 2012 Mayo ruling." The Ivory Leaguers in this Ivory Tower need to be told bluntly that they've created this mess, and so they need to fix it/clean it up. Again, a very faint hope that they will do anything other than to make this situation worse.
Posted by: EG | March 10, 2014 at 07:03 AM
I think I agree. In my mind, the worst case is not that the SCOTUS ruling is so restrictive it makes getting software patents virtually impossible. If that were to happen, I suspect Congress could be persuaded to amend 101 to explicitly permit at least some software patents.
The worst case would be another Mayo-like ruling that provides a test so vague and subjective that it could be used to the same effect, and has us all dealing with trying to apply it in practice for the next few years.
Posted by: Mike Borella | March 10, 2014 at 08:48 AM