In 2014's Alice Corp. v. CLS Bank Int'l case, Justice Thomas famously wrote, "we need not labor to delimit the precise contours of the 'abstract ideas' category in this case."  Instead, he found the claims of patentee Alice to have "no meaningful distinction" under 35 U.S.C. § 101 from those the Court previously found ineligible in Bilski v. Kappos.

Since then, the abstract idea inquiry has been applied by the Federal Circuit over 100 times in various ways.  Initially, there was some hope that this category may be limited to financial transactions or certain types of computer-implemented business methods.  Not so.  Recently, network-controlled electric car charging stations and garage door openers were found to be abstract in ChargePoint, Inc. v. SemaConnect, Inc. and The Chamberlain Group, Inc. v. Techtronic Industries Co., respectively.  These inventions join a long list of those found too abstract for patenting, including claims for displaying advertisements in before content in a video distribution service, dynamic tabs on a user interface, administration of digital images, monitoring of an electrical power grid, and computer virus screening just to name a few.

Even with all of these appellate opinions, the legal definition of "abstract idea" remains remarkably elusive.  Notably, in Interval Licensing LLC v. AOL, Inc., Judge Plager of the Federal Circuit wrote, "a search for a definition of 'abstract ideas' in the cases on § 101 from the Supreme Court, as well as from this court, reveals that there is no single, succinct, usable definition anywhere available."  Last year, the U.S. Patent and Trademark Office (USPTO) published patent eligibility examination guidance that defined three classes of abstract idea (mental processes, mathematical concepts, and methods of organizing human activity).  This demonstrates that the USPTO is struggling to have its examining corps of thousands apply Alice consistently.  But the courts have not adopted the USPTO's view of the world, and the Federal Circuit has explicitly stated that it is not bound by the guidance.

Denying one a property right or taking one's granted property right away based on such ill-defined criteria seems suspicious if not straight from the mind of a genetically-engineering offspring of George Orwell, Franz Kafka, and Terry Gilliam.  And yet, here we are.

The Google dictionary definition of "abstract" is "existing in thought or as an idea but not having a physical or concrete existence."  But one cannot square such a definition with the case law.  Some of the claims in Alice were found abstract despite reciting physical, tangible computer hardware.  Indeed, pure software written in code is stored on a physical tangible medium, such as a disk drive, flash memory, CDROM, etc.  While thinking about software is abstract, the code and equipment that makes your computer display this article -- or do anything else for that matter –– is not.  The dictionary definition does not help us decode how the courts are applying § 101.

Instead, the abstract idea is an elaborate legal fiction that has little to do with the layperson's understanding of what might or might not be abstract.  Without being grounded in statute or plain meaning, it is not hard to see why patentees, patent attorneys, the USPTO, and federal judges remain unable to agree on whether certain types of inventions are eligible for patenting.  And this gray area covers many types of inventions that are not only at the bleeding edge of research, but also crucial to the U.S. economy and competitiveness.  Artificial intelligence, machine learning, data science, and quantum computing innovations are more difficult to patent due to Alice and its progeny.[1]

A Concise Definition

Can one synthesize the aforementioned 100-plus Federal Circuit decisions applying the abstract idea test and come to a more clear understanding of what this category entails?  A daunting task no doubt, but one that I will attempt to address (with a handful of caveats, of course).

Particularly, in view of post-Alice Federal Circuit case law:

a claim is not directed to abstract idea if it includes at least one element that is a specific, technical improvement that is not found in the prior art.

Let's break this down by stating the converse -- a claim is directed to an abstract idea if all of its elements are non-specific, address a non-technical problem, or are known in the art.  Thus, in order for a claim to be successful under § 101, it should have three qualities:  specificity, a technical problem that it solves, and some degree of novelty.  I will address each of these notions in turn.

Specificity

In ruling the invention in McRO, Inc. v. Bandai Namco Games America Inc. eligible, the Federal Circuit used the word "specific" or "specifically" at least 15 times when referring to the claims.  Similar usage of "specific" as a root word can be found in Enfish LLC v. Microsoft Corp., Trading Techs Int'l v. CQG, Inc., and SRI Int'l Inc. v. Cisco Systems Inc. -- in each of these cases, claims were ruled eligible.  In contrast, the Court in Chamberlain lamented the lack of specificity of the claims under review, which -- as noted above -- were found ineligible.

An aspect of specificity is whether the claims recite an outcome or how to achieve that outcome.  Purely outcome-oriented claims often appear functional, indicating what the invention provides rather than how the invention provides it.  In Electric Power Group, LLC v. Alstom S.A., the Court found the claims therein ineligible due to their "defining a desirable information-based result and not [being] limited to inventive means of achieving the result."  Likewise, in Two-Way Media Ltd. v. Comcast Cable Communications, LLC, the Court criticized one of the claim's "result-based functional language."  Particularly, "[t]he claim requires the functional results of converting, routing, controlling, monitoring, and accumulating records, but does not sufficiently describe how to achieve these results in a non-abstract way."  It is not hard to find more examples of claims that have fallen due to lacking specificity regarding the steps needed to carry out the invention.

Indeed, there are five types of specificities that can appear in a claim.[2]  Ideally for § 101 purposes, a claim recites a specific function of a specific element that does something in a specific way to obtain a specific result and thus solve a specific problem.  Not all five specificities need be present in each claim, but the more of these specificities there are, the more likely the claim will survive an eligibility challenge.  Conversely, the fewer specificities in the claim, the more likely it will be found abstract.

If you are thinking that specificity of claim language should be handled by the definiteness, enablement, and written description requirements of § 112 you would not be wrong.  But one of the odd aspects of post-Alice case law is how the § 101 test has evolved to take on characteristics from the § 112 tests.  This evolution has not gone unnoticed at the Federal Circuit.  In American Axle & Mfg. v. Neapco Holdings LLC, Judge Moore wrote in dissent that the majority had applied an enablement test in the guise of eligibility.[3]

Addressing a Technical Problem

But claims require more than just specificity to be eligible.  There are numerous examples of narrow, specific claims that fell to a § 101 challenge, including those of Trading Techs. Int'l, Inc. v. IBG LLC ("[t]he claims are focused on providing information to traders in a way that helps them process information more quickly, not on improving computers or technology"),[4] Versata Development Group, Inc. v. SAP America, Inc. (a claim "akin to creating organizational management charts" fails to "solve a technical problem using a technical solution"), Shortridge v. Foundation Construction Payroll Service, LLC ("generating certified payroll records is an abstract business practice"), and SAP America, Inc. v. InvestPic, LLC ("Under the principles developed in interpreting § 101, patent law does not protect [an advance in mathematical techniques in finance], without more, no matter how groundbreaking the advance").  Thus, specificity will only get you so far and is not the complete answer to the § 101 riddle.

But Trading Techs, Versata, Shortridge, and SAP America all exhibit an unfavorable characteristic -- they are more directed to solving business problems rather than technical problems.  While there is no statutory notion in the U.S. that patents only protect solutions to technical problems, the America Invents Act did introduce the Covered Business Method Review process.  This, along with Alice, may have resulted in the Federal Circuit feeling like it has license to look askance at business methods, with the district courts and USPTO following.

What exactly is a difference between solving a technical problem versus a business problem?  For guidance, we can look to Europe and other jurisdictions where such a concept has been in place for years.  Along these lines of thinking, inventions exist on a spectrum from those addressing technical problems by improving the operation of machines (e.g., engine control, error correction coding of signals, multiprocessing), to those addressing the operation of a business (e.g., derivative trading, hedging risk analysis, ecommerce, and retail schemes).  The former is much more likely to be patentable than the latter.  In the middle, we have a gray area of inventions that improve human performance or life in general (e.g., film animation, desktop communication applications, video game avatar controls).  These inventions could go either way.

Indeed, the Federal Circuit has even found inventions that are claimed rather broadly to be eligible due to their technical nature.  In SRI, claims addressing a hierarchical network monitoring system were found rooted in computer technology in order to solve a specific problem arising in the realm of network computing.  Likewise, claims in Koninklijke KPN N.V. v. Gemalto M2M GmbH were also found eligible "because they are directed to a non-abstract improvement in an existing technological process."

Thales Visionix Inc. v. U.S. is a notable case in this category, where remarkably broad claims were found to be eligible because the accompanying specification explained in detail the technical improvement they provided.  But one should not conclude that specificity can be replaced by a technical improvement.  Claims with both fare better, and Thales may be an outlier.  A lengthy explanation of an invention's technical character in the specification (when argued properly) probably cannot hurt claims, but it will not guarantee success if the claims are outcome-oriented.

The USPTO's Patent Trial and Appeal Board (PTAB) seems to be taking the European approach.  Recently, the PTAB published a number of informational decisions applying the USPTO's latest examination guidance.  The outcomes of these cases break along the technical problem to a technical solution line.  An invention directed to optimizing the signal to noise ratio from reception coils of an magnetic resonance tomograph was found to meet the § 101 requirements, as was an invention focused on transforming the coordinate system of catheter navigation to the coordinate system of an external model or external image data.  On the other hand, software for selecting a golf club with a suitable loft angle and software for creating a life-cycle workflow for a project were found ineligible due to their addressing non-technical problems.

Nonetheless, the contours of the business method are somewhat malleable and in the eyes of the beholder.  There are instances of the Federal Circuit finding claims to be eligible where those claims are ostensibly business-oriented.  In Trading Techs Int'l v. CQG, Inc., the Court concluded that claims directed to a securities trading user interface was sufficiently technical because it solved existing problems in such interfaces.  In DDR Holdings, LLC v. Hotels.com L.P., the Court found that the claimed solution of "generating a composite web page that combines certain visual elements of a host website with content of a third-party merchant" to be technical and "necessarily rooted in computer technology," despite arguably addressing the business-oriented problem of retaining or increasing web site traffic.

Not in the Prior Art

Patent eligibility under § 101 was not intended to be a prior art test, but it has become one.  Despite Judge Moore's criticism of the Court's combining § 112 considerations into § 101, her opinion in Berkheimer v. HP Inc. introduces factual considerations of prior art under §§ 102 and 103 into part two of the Alice test (the search for an inventive concept where one must consider whether claim elements are well-understood, routine, and conventional).  But the cross-contamination of eligibility and prior art began before Berkheimer.  Indeed, comparisons between claimed inventions and prior art are found in a vast number of § 101 decisions including the aforementioned KPN, Chamberlain, Enfish, and McRO cases, as well as Cellspin Soft, Inc. v. Fitbit, Inc., Data Engine Technologies LLC v. Google LLC, Core Wireless Licensing S.A.R.L. v. LG Electronics, Inc., and Internet Patents Corp. v. Active Network, Inc., just to name a few.

The exact nature of this de facto prior art analysis under § 101 is unclear.  Not only does it impact the second part of the test but also the first part.  For example, in Chamberlain, the Court found the claims abstract after noting that the "only described difference between the prior art movable barrier operator systems and the claimed movable barrier operator system is that the status information about the system is communicated wirelessly" and that wireless communication is an abstract concept.  Rather than analyzing claims as a whole under part one as was set forth in Enfish and McRO, recent Federal Circuit cases appear to be dividing claims into elements within the prior art and those that are not.  The Court then only considers whether the non-prior-art elements are specific and technical enough to be more than an abstract idea.

But what if all of your claim elements are in the prior art, but your claims recite a combination that is not?  As noted, even though claims are supposed to be analyzed as a whole under part one of Alice and as an ordered combination under part two, these analyses get short shrift in practice.  In cases like Chamberlain and InvestPic, the Federal Circuit provided conclusory reasoning by using variations of the phrase "when considered as an ordered combination, the components of each claim add nothing that is not already present when the steps are considered separately."  This language is derived from Mayo, and similar wording regarding ordered combinations can be found in numerous district court and PTAB decisions.  Bottom line -- relying on claims being viewed as a whole or as an ordered combination is risky.

Analysis

To be clear, the definition I propose is not perfect, and is merely an attempt to fit a line through a messy set of data points.  Still, following this approach can provide a sense of how likely it is for your invention to avoid § 101 issues, or whether drafted claims should be edited or amended to become more clearly eligible.  While some of these factors might be able to be downplayed in certain situations, the more specific, technically-oriented, and novel your claimed invention, the higher the probability of a desirable outcome.

Notably, this definition is not the two-part Alice test.  There is no "directed to" or "inventive concept" inquiries.  In practice, the Alice test often effectively has only one part -- whether the claim is directed to a judicial exception.  In conducting this inquiry for the abstract idea category, one considers the claim's specificity, technical character, and whether it recites elements in the prior art.  Therefore, I posit that the definition proposed is a more coherent representation of the process actually carried out than the disparate articulations of Alice in the last five-and-a-half years of Federal Circuit case law.

[1] Medical diagnostics and genetic engineering are other high-value fields that suffer from the Supreme Court's rulings, though inventions in these areas are generally subject to the natural phenomenon or law of nature exceptions to eligibility, and are thus controlled by Mayo Collaborative Services v. Prometheus Labs., Inc. and Association for Molecular Pathology v. Myriad Genetics rather than just Alice.

[2] These five specificities are different but in the same spirit as those introduced by Judge Lourie in Vanda Pharmaceuticals v. West-Ward Pharmaceuticals.  In that case, he wrote "[a]t bottom, the claims here are directed to a specific method of treatment for specific patients using a specific compound at specific doses to achieve a specific outcome."

[3] Admittedly, the judicial exception in American Axle was a law of nature, but Judge Moore's general reasoning holds.

[4] There are multiple Federal Circuit cases with this heading, and the one decided April 30, 2019 is referred to here.