Benevolent despot or tyrant? Alice v CLS Bank five years on

Benevolent despot or tyrant? Alice v CLS Bank five years on

The law on patent-eligible subject matter has been transformed since the Supreme Court’s Alice decision with profound effects for patent owners, the courts and the USPTO

June 2019 will mark the fifth anniversary of Alice Corp Pty Ltd v CLS Bank Int’l. In Alice, the US Supreme Court announced that patent-eligible subject matter is determined based on a two-step test. First, one determines whether the claim is directed to a “judicial exception” to Section 101 of the Patent Statute, specifically a law of nature, an abstract idea or a natural phenomenon. If so, then one looks for something more, to wit, an ‘inventive concept’ that transforms the otherwise ineligible claim into a patent-eligible application of the judicial exception.

The application of this test under Queen Alice’s reign has dramatically altered the patent landscape. More than 970 patents have been invalidated by the courts and the USPTO’s PTAB, while more than 60,000 patent applications have been abandoned before the USPTO following rejections for patent-ineligible subject matter. Patents and portfolios in many fields – particularly software and biotechnology – have declined in value or simply become unsaleable at any price.

Your interpretation of these facts depends on your point of view. If you are a rights holder or licensor who depends on the certainty of patent rights, then the numbers are merely a grim confirmation that bad patents and bad science make for bad law. The consistency that Alice brings to litigation is, at best, the general certainty that a patent on just about any kind of technology can be subject to a motion to dismiss for ineligible subject matter – and that nearly 60% of such attacks succeed in the district courts and are then affirmed more than 90% of the time on appeal to the Federal Circuit. You agree with the assessment that the Alice test is a fancy ‘I know it when I see it’ shorthand for judges to use to decide whether patent claims have so-called ‘inventive merit’ – an approach that one judge of the Federal Circuit has described as providing “no discernable boundaries for decision-making in specific cases, resulting in an incoherent legal rule that led to arbitrary outcomes”. This observation echoes that of Judge Rich, the author of the Patent Act 1952, who stated that: “Nowhere in the entire act is there any reference to a requirement of ‘invention’ and the drafters did this deliberately in an effort to free the law and lawyers from bondage to that old and meaningless term.” Nonetheless, with Alice the hunt for an invention is back, only this time it is typically a proxy for invalidating patents for obviousness, lack of written description or enablement without the costly need for evidence. Thus, you view the Federal Circuit’s decisions as inconsistent at best and based on an arbitrary division between claims using a computer as a tool (generally ineligible) and claims for improving the computer itself (generally eligible). Your experience before the USPTO has been similar; until recently, patent examination of your company’s software and/or biotech inventions have been dominated by Section 101 rejections, increasing your prosecution costs and the time to issuance. The value of your portfolio has dropped – possibly to zero – and some licensees have aggressively attempted to renegotiate their payments.

On the other hand, if you are a company that is a target of patent assertions, then these numbers are cause for celebration: Queen Alice has made the world a better place by reducing your exposure to both the meritless claims of existing patents and – better yet – the possible universe of future patents that could have been used against you. If you have had to forgo patent protection of your own – by not filing, abandoning a few applications here and there, or just dealing with increased prosecution costs – it has been a small price to pay for the increased certainty of outcomes. Queen Alice has been benevolent and wise, and objections to her rules are merely Talmudic carpings of patent maximalists and nostalgic, pernickety patent attorneys.

In this review we will unpack these trends and provide further insights into the problems and prospects of Alice’s reign.

Table 1. Summary of Section 101 litigation: July 2014 - April 2019



Valid patent claims

Invalid patent claims

% invalid

Federal court decisions





Federal Circuit





District courts










Motions on pleadings





* Total patents is less than the sum of the invalid and valid columns since some patents were found to be both valid and invalid

Figure 1. Federal courts, patent eligibility decisions and trends: July 2014 - April 2019

Queen Alice at court

Table 1 summarises the litigation landscape of Section 101-related cases from July 2014 to April 2019. In that period, there were 805 federal court decisions based on Section 101 motions. Overall, those motions were successful about 63% of the time. In particular, motions on the pleadings (including motions to dismiss) were granted 59.1% of the time. The Federal Circuit has been particularly vigorous in applying the Alice test, finding patent-eligible subject matter in fewer than 15% of appeals – just 20 times in 150 decisions. In the past five years, 746 unique patents have been invalidated in whole or in part by the federal courts.

Looking at the consolidated information over time fleshes the story out even further.

Figure 1 shows the data from Table 1 broken down by quarter. Over time the percentage of decisions finding patents invalid has been slowly declining and levelling out on a quarterly basis as shown by the four-quarter rolling average of the percentage of decisions finding patents ineligible, which now hovers around 51% per quarter. We attribute this behaviour to a number of factors, most significantly the reduction of patent cases with claims that were weak to begin with (ie, likely obvious, lacking in enablement or other technical problems). This trend suggests that Section 101 has been effectively used as a proxy by the courts to invalidate these kinds of patents. Thus, over time, patentees, as well as contingency fee-based litigation firms, have become more selective in the patents that they assert.

When it comes to appeals, the Federal Circuit has been steadily assertive in upholding lower court decisions of ineligibility. Overall, the Federal Circuit has affirmed 87% of lower court decisions invalidating patents. Patentees have faced a major uphill battle over the last five years: they have lost before the district court about 59% of the time on early dispositive motions and the Federal Circuit has affirmed 87% of such decisions. Indeed, the Federal Circuit often affirms such appeals without issuing any written opinion at all using Rule 36 in 35% of its decisions – an outcome that is heavily criticised because it precludes the patentee from seeking Supreme Court review on what is said to be a pure question of law. When it comes to the USPTO’s PTAB, the Federal Circuit has been even more accommodating to patent defendants: it has affirmed 100% of PTAB covered business method reviews that invalidated the patents in suit, as well as 100% of appeals from ex parte decisions affirming Section 101 rejections of patent claims in pending applications.

Queen Alice’s regime started with the invalidation of business method patents and these types of patents tend to draw the most criticism – even though the Supreme Court in Bilski made it clear that there is no categorical exclusion of business methods from the Patent Act. Nonetheless, the Federal Circuit has found every business method patent that it has considered since Alice ineligible. Yet this is not the whole story. In fact, the vast majority of patents invalidated since Alice have been software patents of various types. Figure 2 illustrates the breakdown of patents that have been litigated under Section 101 after Alice.

It would be a mistake to think that patents outside of the core areas of software and communications are less likely to be attacked by an aggressive defendant. The 50 patents in the ‘other’ category come from fields as diverse as metal working, gas separation, water treatment, oil drilling, refrigeration, plastics and sewing; and – as the data in Figure 2 shows – the majority (57%) have been invalidated.

Table 2 shows the breakdown of the overall data according to the type of motions.

Here we see that types of motion for judgment on the pleadings are most successful overall by a wide margin over motions to dismiss. This suggests that patent defendants are best served by filing their motions at this stage.

Finally, the PTAB has been the most aggressive enforcer of Queen Alice’s rule. Table 3 summarises the outcome of covered business method reviews and post-grant reviews that raise Section 101 defences.

Figure 2. Patent eligiblity outcomes for patents by technology type: July 2014 - April 2019

Table 2. Summary of motion outcomes and appeals: July 2014 - April 2019


Total decisions

Not invalid


% invalid






Motion for JOP





Motion to dismiss










Post-trial motion










Fed Cir






























Appeal-prelim Inj










Appeal-post-trial motion





Table 3. Summary of Section 101 outcomes in PTAB covered business method reviews
July 2014 - April 2019



Valid patent claims

Invalid patent claims

% invalid

PTAB total





PTAB CBM petitions





PTAB CBM final decisions





PTAB PGR petition





Analysis of Federal Circuit case law

What is driving these outcomes is a set of interrelated doctrines and judicial biases that form something of an inescapable trap. First and foremost is the Federal Circuit’s reduction of the Alice test into a set of shorthand rules for what constitutes an abstract idea. In Step 1 of the Alice test, when it comes to computer-implemented inventions the court has essentially instituted the rule that claims that merely use a computer as a so-called ‘tool’ for some other purpose are presumed to be directed to an abstract idea – regardless of whether the claims are directed to a judicial exception. On the other hand, claims that seem to improve the performance or operation of the computer itself or other physical technology, on the other hand, are not abstract. This approach has taken what the Supreme Court used as a single example of possible basis for an inventive concept in Step 2 (improvements in computer performance) and turned it into a fixed marker in Step 1. In essence, whatever function is being performed by the computer is deemed to be an abstract idea per se and so the court disregards the presence of any conventional physical components as merely being the necessary or obvious means by which the function is to be performed.

Within the zone of using a computer as a tool, the Federal Circuit has developed several broad categories of abstract ideas. The most significant of these is collecting, analysing and displaying information from Electric Power Group v Alston – a description that is handily broad enough to encompass the majority of computer-implemented inventions and which gives the courts enormous latitude in deciding whether to categorise an invention as abstract. In short, the Federal Circuit has taken what is the core virtue of computers – a programmable general-purpose information processing capability using standardised hardware – and used it as a signal of patent ineligibility. Other categories of presumptively ineligible abstract ideas are:

  • gaming (electronic or otherwise);
  • economic activities; and
  • processes that a human could hypothetically perform, even if such performance was never contemplated by the inventors or even remotely likely in real life.

This facade covers what is actually going on with the Federal Circuit and the lower courts: the use of Section 101 as a proxy for invaliding patents that the courts would otherwise consider obvious, lacking enablement or written description – without the costs in time and resources required by current doctrines. When confronted with claims that use computers to automate known or obvious processes, a court finds such claims to be directed to an abstract idea in Step 1 and then in Step 2 finds that there is nothing that transforms the claims since all of the elements use conventional computer elements (eg, processors, memory, displays, networking components and so on): “It is well settled, though, that automating conventional activities using generic technology does not amount to an inventive concept” (see LendingTree v Zillow, ineligible claims for internet-based loan applications). For example, in Fairwarning IP LLC v Iatric Systems, Inc, the Federal Circuit invalidated claims to systems and methods of detecting fraudulent access of medical information on the grounds that “FairWarning’s claims merely implement an old practice in a new environment”. (See also In re Salwan (ineligible patent application claims for electronic medical record management are “little more than the automation of a ‘method of organizing human activity’ with respect to medical information”); Tranxition, Inc v Lenovo (United States) (ineligible claims for automated migration of computer configuration information); and OIP Technologies, Inc v, Inc (ineligible claims for automated price optimisation)).

Alternatively, if the court faces an invention that cannot be reduced to mere automation, it can use Section 101 as a proxy for lack of enablement or written description. In these cases, the court typically asserts that the claims or the specification do not describe how the invention performs the claimed functions. The court took this approach to the extreme in Glasswall Solutions Limited v Clearswift Ltd, taking each step of the claim individually and finding that it did not describe how the step was performed:

The claims at issue in both patents do not purport to claim how the invention receives an electronic file, how it determines the file type, how it determines allowable content, how it extracts all the allowable data, how it creates a substitute file, how it parses the content according to predetermined rules into allowable and nonconforming data, or how it determines authorization to receive the nonconforming data.

(See also Voit Technologies, LLC v Del-Ton, Inc (Claims ineligible where “it is unclear how the different compression format claim limitations actually achieve the alleged improvements”.); Vehicle Intelligence v Mercedes-Benz USA, LLC (“the claims do not specify what screening should be done or how the expert system would perform such screening… how to select the tests… how the expert system works… or how such systems can be portioned out.”)).

In these cases, the court invariably criticises the claims as being purely functional in nature: “Instead, the claims are framed in wholly functional terms, with no indication that any of these steps are implemented in anything but a conventional way.” Of course, this kind of analysis can be applied to just about any method claim, since what is ‘functional’ and what is ‘implementation’ depends on the engineer’s level of skill. This is why in a highly skilled field such as software engineering, purely functional descriptions are often sufficient for enablement, as a skilled programmer can readily implement the function in any variety of different ways. Further, if the real failure here is a lack of description of how the invention operates, that is a problem in enablement or written description, not subject-matter eligibility.

To argue that claims to software-based inventions are ineligible for being functional reveals a failure to understand that inventions in software are by their nature functional. The central premise of computer science and engineering is the creation of a general purpose hardware system that can be programmed by software to perform new and specific functions. Thus, in many cases the invention is in the functional behaviour itself, which by design employs the conventional (ie, existing services and capabilities of the computer) to achieve the results. No one expects a patent claim that incorporates an electric motor to limit the claim to how the motor operates – thus one should not require a patent claim that incorporates a specific computer function (that is fully enabled in the specification) to recite the hardware implementation of the function. Nonetheless, that has become the rule under Alice and her subjects should be loath to ignore it.

Queen Alice has likewise applied her rules to biotech inventions. In Mayo the Supreme Court defined a ‘law of nature’ as any natural correlation that occurs without human intervention. This definition has no scientific basis – indeed no scientist would consider every natural correlation to be a law of nature – since as every student of statistics knows, correlation is not causation. Nonetheless, Queen Alice has again prevailed. The result has been more than 80 biotech-related patents – particularly those related to diagnostic methods – being invalidated. Here, however, the Federal Circuit judges – even those who see no difficulty in identifying abstract ideas – have spoken out that the Supreme Court’s approach is problematic. Judge Dyk has written that Mayo may screen out “useful new diagnostic and therapeutic methods”. Judge Lourie likewise argues that “it is unsound to have a rule that takes inventions of this nature out of the realm of patent-eligibility on grounds that they only claim a natural phenomenon plus conventional steps”. Nonetheless, these judges make clear that they are “bound by the sweeping language” of Mayo.


One of the most significant issues in the past five years is the degree to which patent eligibility is a mixed question of law and fact – and thus under what circumstances a patentee can defeat an early dispositive motion (which are typically based on pure questions of law) on the grounds that there are material facts in dispute. While the Supreme Court has treated eligibility as a question of law, it has never actually addressed this issue on the merits. Further, by introducing the consideration as to whether the claim elements are “routine, well-understood and conventional” in Step 2 of the Alice test, as well as statements that Section 101 can involve considerations similar to Sections 102 and 103, the court opened the door to the role of facts in patent eligibility.

In two decisions in 2018, Berkheimer v HP Inc and Aatrix Software, Inc v Green Shades Software, Inc, the Federal Circuit clarified that patent eligibility is a mixed question of law and fact. In Berkheimer, it held that: “The question of whether a claim element or combination of elements is well-understood, routine and conventional to a skilled artisan in the relevant field is a question of fact.” Thus, where there is a genuine issue as to whether a claim element or combination of elements is well understood, routine and conventional, summary judgment of patent ineligibility should be denied. Similarly, in Aatrix, the court held that well-pled allegations that certain claim elements contain an inventive concept and are not routine and conventional can preclude a finding of ineligibility on motion to dismiss. En banc review of Berkheimer was denied, which revealed the continuing difficulties of applying the Alice test. While Judges Moore, Dyk, O’Malley, Taranto and Stoll concurred in the denial and called the result “unremarkable”, Judge Reyna’s dissent argued that Aatrix and Berkheimer altered “in a significant and fundamental manner” the “single most consistent factor” of eligibility analysis: that Section 101 is a question of law. Not coincidentally, Reyna has both the most decisions (34) and a very high percentage (92%) of decisions finding ineligible subject matter.

Most notable about the en banc denial, however, was the concurrence by Judges Lourie and Newman that explicitly called for clarification by “higher authority, perhaps by Congress”. Lourie and Newman have often been considered on opposite extremes in interpreting Section 101 jurisprudence; however, in this instance they wrote together to note that the present analysis prevents patenting of “meritorious inventions… in a way that revolutionized prenatal care”. Likewise, with respect to abstract ideas, they stated that the need for a second step is analytically problematic:

If a method is entirely abstract, is it no less abstract because it contains an inventive step?… Step two’s prohibition on identifying something more from ‘computer functions that are well-understood, routine, conventional activities previously known to the industry’ is essentially a §§ 102 and 103 inquiry. Section 101 does not need a two-step analysis to determine whether an idea is abstract.

As of the time of writing, Berkheimer is pending certiorari before the Supreme Court, with the court recently requesting the solicitor general’s opinion as to whether to grant the petition.

These decisions were viewed positively by patent owners as offering a way past the gauntlet of early dispositive motions, as discussed above. However, the outcomes have been mixed. Of the 50 cases in which Berkheimer and/or Aatrix have been cited by the district court in a motion to dismiss or judgment on the pleadings, 25 (50%) have still granted these motions. Of 18 cases decided on motion for summary judgment, seven (39%) district courts granted the motion invalidating the patents in suit, while 11 (61%) found the claims not invalid at the summary judgment stage. At the Federal Circuit, of 11 cases citing Berkheimer, 10 (91%) affirmed invalidity. In short, Berkheimer has not reduced the risk of early dispositive motions as anticipated.

The actual impact of Berkheimer at the Federal Circuit may be more muted in part because of the flexible way in which an abstract idea may be defined. For example, the Federal Circuit has interpreted limitations that provide specificity to the abstract idea as merely a “more specific” abstract idea, rather than calling for evaluation of whether such details are sufficient under Step 2 to transform the abstract idea into a practical application. With regard to patent prosecution, the USPTO has issued a memorandum requiring examiners in a prima facie rejection to demonstrate, with evidence, that additional limitations in a claim were well understood, routine and conventional. Nonetheless, many examiners skirt that requirement by shoehorning most of the claim into their definition of the abstract idea. This leaves only a processor or other basic element for consideration under Step 2B, resulting in the foregone conclusion that there is no inventive concept.

The difficulties with the Alice test and the division between the Federal Circuit judges on the basic question of whether patent eligibility is a pure question of law are certainly known by the Supreme Court. There have been 39 petitions for certiorari from the Federal Circuit by patentees whose patents have been invalidated and all have been denied. It appears that the Supreme Court continues to believe that the Federal Circuit will find a way to make the unworkable test workable after all.

Table 4. Percentage of office actions with Section 101 rejection, by primary class and year

Table 4

Queen Alice in the office

Over the past five years, the USPTO has attempted to digest the holdings of the federal courts and summarise them for use by the patent examiners working in the trenches and as guidance for patent applicants. Between 2014 and 2018, the office issued 12 memoranda on patent eligibility, as well as over 40 different example claims to help examiners to formulate a proper eligibility rejection. Nonetheless, this guidance failed to bring much consistency to eligibility analysis – other than the near certainty that patent applications in particular technology areas of the USPTO would receive Section 101 rejections.

For example, in the July 2015 guidance, the USPTO so broadly characterised the holdings of various federal court decisions (eg, “obtaining and comparing intangible data” or “comparing new and stored information and using rules to identify options”) that virtually any invention implemented with software could be characterised as constituting a mere abstract idea, should the examiner wish to do so. Thus in some art units – especially those dealing with finance, banking, healthcare and business operations – more than 90% of applications were receiving these rejections.

More specifically, we collected the dataset of every office action issued between 2014 and 2018 and evaluated the percentages of office actions that contained rejections for patent-ineligible subject matter, with respect to the technology. This is summarised in Table 4.

As shown, certain technology areas had very high and consistent rates of Section 101 rejections. These were primarily patents examined in Technology Center 3600, which handles business methods. Interestingly, the peak for such rejections appears to have been in 2016, while for other areas the rates of such rejections climbed in 2017 and 2018. Further, our research revealed that some examiners had issued Section 101 rejections in 100% of the applications that they examined and had not granted any patent applications between 2016 and 2018.

The appointment of Andrei Iancu as director of the USPTO resulted in the most significant change to date in the office’s handling of Section 101. In many speeches during 2018, Iancu repeatedly expressed concern about uncertainty in the present law and the problems that such uncertainty creates for patent examiners, practitioners and patent owners. He articulated that in his view that the core logic of Section 101, as created by the Judge Giles Rich, was to avoid “the clear commingling of distinct statutory provisions which are conceptually unrelated, namely, those pertaining to the categories of invention in §101 which may be patentable, and to the conditions for patentability demanded by the statute”. Quoting Rich, the author of the Patent Act 1952, Iancu stated that “when we deal with §101, the sole question… is whether the invention falls into a named category, not whether it is patentable”. Thus, the critical task for the USPTO is to provide examiners with an objective way to determine which claims are within the categories of eligible subject matter – and which claims are not. As Iancu noted, just as the categories of eligible subject matter have been articulated by the statute, so too must the judicial exceptions to it be clearly categorised, specifically with regard to so-called ‘abstract ideas’.

A change in tone at USPTO

The USPTO issued its most recent examination guidance in January 2019, which distinctly changed the tone of Section 101 analysis, making it more conducive for examiners to find patent claims eligible subject matter. As a reminder, the first step of the Alice test is to determine whether a claim is directed to a judicial exception (eg, an abstract idea) and in the second step to consider whether there is an inventive concept.

The 2019 Guidance divided the first step into two prongs. In the first, the examiner decides whether a claim recites the judicial exception idea directly or indirectly. Here, the 2019 Guidance moves away from providing an expansive list of numerous sub-categories of abstract ideas to be mined by examiners, as had been the case previously. Instead, it returned to using the three basic categories of abstract ideas previously identified by the courts:

  • mathematical concepts;
  • methods of organising human activity; and
  • mental processes.

Next, if a claim is found to recite an abstract idea, then in the new second prong the examiner must decide “whether the claim as a whole integrates the recited judicial exception into a practical application of the exception” before considering whether there is an inventive concept. If a practical application is found, then the claim is not directed to the abstract idea and there is thus no need to search for the inventive concept.

This practical application test is the USPTO’s attempt to synthesise how the Supreme Court has approached patent eligibility historically. The 2019 Guidance states that claim elements that indicate a practical application are those that reflect an improvement in the functioning of the computer, apply the judicial exception in a particular manner or result in a particular transformation – as well as other considerations. The overall approach of the guidance is to shift the focus of the inquiry from finding a specific abstract idea to finding a practical application of that idea.

Crucially, the 2019 Guidance states that whether a claim element is well understood, routine or conventional has no bearing on whether there is a practical application. This is an important caveat because many examiners routinely lumped all potentially novel and non-obvious elements into the abstract idea under Step 1 of Alice and then found no inventive concept under Step 2 since everything else was, by assumption, conventional.

The real benefit of the 2019 Guidance came not in its text per se but in the overall approach, which orients the examiner towards finding eligible subject matter, rather than towards finding a basis for a rejection. While the 2019 Guidance may appear little more than a different restatement of court cases, the newly issued example claims (see demonstrate best how the office expects the guidance to be implemented. These examples show that the USPTO expects most claims – even those that are found to recite an abstract idea – to still have sufficient limitations to recite a practical application and therefore be patent eligible. The difference between an ineligible claim and an eligible one can be as simple as that between claiming “determining a X by processor” (ineligible) and “determining X by a processor using element Z to determine on value V” (eligible). The key is that the latter claim limitation provides the practical application by limiting the determination of X to a particular way that can only be performed by the computer itself, rather than by a human being. Based on the behaviour of patent examiners as of early March 2019 (eg, in the number of cases summarily dropping many longstanding eligibility rejections), the 2019 Guidance is shaping up to be the most significant pro-eligibility development since Alice.

That said, there is some concern within the patent community that the 2019 Guidance conflicts with various decisions of the Federal Circuit and thus patents that issue based on it may nevertheless be invalidated by the courts, particularly in the life sciences. Several Federal Circuit decisions suggest that a practical application alone may be insufficient to avoid patent ineligibility. For example, in Ariosa Diagnostics, Inc v Sequenom, Inc, one of the claims found patent ineligible was not simply drawn to detecting a paternally inherited nucleic acid but rather to a “method for performing a prenatal diagnosis on a maternal blood sample”. This appears to be a practical application of detecting paternally inherited nucleic acid, albeit still of broad scope. Similarly, the dependent claims were found ineligible even though they required specific methods of application and specific probes – details that would appear to satisfy the additional elements aspects set out previously. Further, in Athena Diagnostics, Inc v Mayo Collaborative Servnces LLC, the Federal Circuit held that an application of a correlation between a particular antibody and a disease (mysethenia gravis) to “diagnos[e] neurotransmission or developmental disorders related to [MuSK] in a mammal” was patent ineligible. Here too, some claims recited additional elements such as specific reagents that were integral to the method (labelled MuSK constructs that would bind to the MuSK antibodies being detected) that would appear to satisfy the USPTO’s requirements that an “additional element applies or uses the judicial exception in some other meaningful way” or “an additional element effects a transformation or reduction of a particular article to a different state or thing”, but were insufficient for the Federal Circuit.

The same concern arises with software-related inventions. The Federal Circuit has generally held that claims that recite the use of a computer as a tool to perform a function that humans could perform are ineligible. Some aspects of the 2019 Guidance that allow examiners to find a practical application – such as by limiting the operation of the computer to a very specific implementation – may be viewed by the court as merely adding specificity to the abstract idea, rather than limiting the claim in meaningful manner. It may be several years before a patent issued under the 2019 Guidance comes up to the Federal Circuit for review. Until that time the uncertainty remains.

Legislative efforts

The impact of Alice’s reign has prompted many to call for a legislative solution. Federal Circuit Judges Plager and Lourie have stated that “the law needs clarification by higher authority, perhaps by Congress”, given the “current mess regarding our §101 jurisprudence”. David Kappos, former director of the USPTO, has called for the complete abolition of Section 101 from patent law.

Numerous organisations have set out proposals to amend Section 101. These include the Intellectual Property Owners Association (IPO), the American Intellectual Property Lawyers Association (AIPLA) and the American Bar Association (ABA). The IPO and AIPLA have issued a joint proposal that defines two specific exceptions to patent-eligible subject matter: products or processes that exist naturally and purely mental processes. The proposal expressly eliminates any consideration of novelty, non-obviousness or enablement, as well as any of the requirements of the Alice test, from the eligibility issue. The intent of this approach is to return Section 101 to its original focus on categories of subject matter and eliminate the unworkable aspects of Alice. The language of the IPO-AIPLA joint proposal is as follows:

Eligible Subject Matter.

101 a) Whoever invents or discovers, and claims as an invention, any useful process, machine, manufacture, composition of matter, or any useful improvement thereto, shall be entitled to a patent thereof, subject only to the exceptions, conditions, and, requirements set forth in this Title.

Sole Exceptions to Subject Matter Eligibility

101 b) A claimed invention is ineligible under subsection (a) only if the claimed invention as a whole, exists in nature independently of and prior to any human activity, or is to be performed solely in the human mind.

Sole Eligibility Standard

101 c) The eligibility of a claimed invention under subsections (a) and (b) shall be determined without regard to

(i) the requirements or conditions of sections 102, 103, and 112 of this title;

(ii) whether the claimed invention, or any limitation thereof is conventional, routine, or well-understood;

(iii) the manner in which the claimed invention was made or discovered; or

(iv) whether the claimed invention includes an inventive concept.

Importantly, there is no mention in the IPO-AIPLA proposal of the existing judicial exceptions to Section 101. This was to avoid the Supreme Court interpreting the language as endorsing its jurisprudence by codifying the exceptions themselves and thus incorporating the existing Alice test back into the statute – precisely the opposite of the desired outcome. The IPO-AIPLA proposal has been endorsed by a number of bar associations and other groups, including the IP or patent bar associations of Boston, Chicago, New Jersey, New York and Philadelphia, as well as the National Association of Patent Practitioners.

The ABA’s proposed amendment takes a different approach. First, unlike the IPO-AIPLA, the ABA expressly uses pre-emption of the judicial exceptions as basis for ineligibility. The ABA then adds a further test of a practical application as a way of getting claims out of the ineligible zone – something similar in spirit to the USPTO’s 2019 Guidance. The ABA’s language is as follows:

101. Conditions for patentability: eligible subject matter.

(a) Eligible Subject Matter.- Whoever invents or discovers any useful process, machine, manufacture, or composition of matter, or any useful improvement thereof, shall be entitled to obtain a patent on such invention or discovery, absent a finding that one or more conditions or requirements under this title have not been met.

(b) Exception.- A claim for a useful process, machine, manufacture, or composition of matter, or any useful improvement thereof, may be denied eligibility under this section 101 on the ground that the scope of the exclusive rights under such a claim would preempt the use by others of all practical applications of a law of nature, natural phenomenon, or abstract idea. Patent eligibility under this section shall not be negated when a practical application of a law of nature, natural phenomenon, or abstract idea is the subject matter of the claims upon consideration of those claims as a whole, whereby each and every limitation of the claims shall be fully considered and none ignored. Eligibility under this section 101 shall not be negated based on considerations of patentability as defined in Sections 102, 103 and 112, including whether the claims in whole or in part define an inventive concept.

One concern with the ABA’s approach is that using the concepts of pre-emption and the judicial exceptions invites the Supreme Court to interpret the provision as simply being a codification of the Alice test itself. The court has repeatedly stated that the purpose of the Alice test is to exclude patents that would pre-empt use of “laws of nature, natural phenomena, and abstract ideas”. As of the time of writing, no bar associations or other IP-focused groups have endorsed the ABA’s proposal.

These efforts are attracting attention in Congress. Last year, Representative Massie (R-Ken) introduced HR 6264, which includes a provision to amend Section 101 based on the IPO-AIPLA proposal, and which states that “this amendment effectively abrogates Alice Corp. v. CLS Bank International, 134 S. Ct. 2347 (2014) and its predecessors to ensure that life sciences discoveries, computer software, and similar inventions and discoveries are patentable, and that those patents are enforceable”.

More recently, Senators Coon (D-Del) and Tillis (R-NC), along with Representatives Collins (R-Ga), Johnson (D-Ga) and Stivers (R-Ohio) have introduced a framework for Section 101. The framework follows after several months of meetings with members of the patent community. The framework is as follows (see

  • Keep existing statutory categories of process, machine, manufacture, or composition of matter, or any useful improvement thereof.
  • Eliminate, within the eligibility requirement, that any invention or discovery be both “new and useful.” Instead, simply require that the invention meet existing statutory utility requirements.
  • Define, in a closed list, exclusive categories of statutory subject matter which alone should not be eligible for patent protection. The sole list of exclusions might include the following categories, for example:
    • Fundamental scientific principles;
    • Products that exist solely and exclusively in nature;
    • Pure mathematical formulas;
    • Economic or commercial principles;
    • Mental activities.
  • Create a “practical application” test to ensure that the statutorily ineligible subject matter is construed narrowly.
  • Ensure that simply reciting generic technical language or generic functional language does not salvage an otherwise ineligible claim.
  • Statutorily abrogate judicially created exceptions to patent eligible subject matter in favor of exclusive statutory categories of ineligible subject matter.
  • Make clear that eligibility is determined by considering each and every element of the claim as a whole and without regard to considerations properly addressed by 102, 103 and 112.

Superficially, this appears to be good start towards amending Section 101, incorporating principles found in various proposals. However, the framework raises some serious concerns.

First, the framework states it will “abrogate judicially created exceptions to patent eligible subject matter in favor of exclusive statutory categories of ineligible subject matter”. However, unless the exclusive statutory categories are expressly defined in the statute, then the courts will have to craft their own interpretation of these categories, and as such will clearly return to the existing Federal Circuit and Supreme Court interpretations of these categories – which are the very source of the problems to date.

On the other hand, attempting now to define these categories invites a host of problems. For example, while it sounds reasonable to exclude fundamental scientific principles, the problem is determining whether some claimed principle is really fundamental in a scientific sense. As mentioned above, the Supreme Court’s definition of a law of nature is scientifically invalid and if used as the basis of interpretation here would result in an overly broad scope of ineligible subject matter. Further, whether a scientific principle is fundamental is an empirical question that often takes years, sometimes decades, of experimentation to validate. Indeed, most alleged scientific principles turn out to be wrong in fact once rigorously tested. This is why it often takes many years for a scientist to receive a Nobel Prize for their research. Neither a patent examiner nor a court is equipped to judge whether a scientific principle is fundamental. The other exclusive categories suffer from other types definitional problems.

Second, the practical application test is generally unnecessary and creates yet another potential trap. The framework already includes “existing statutory utility requirements”. If an invention has utility as claimed, then it would presumably have a practical application and a separate test is unnecessary. Further, the USPTO adopted the practical application as a way to save claims that are deemed directed to ineligible subject matter under Alice. However, if the intent of the legislation is to essentially overrule Alice by providing clear and strict boundaries for what is excluded subject matter, then the practical application test is again not needed.

Third, and perhaps most problematic, is the statement that “simply reciting generic technical language or generic functional language does not salvage an otherwise ineligible claim”. This statement appears to be driven by those who seek to forestall or perhaps preclude patents on software inventions, which typically use terms such as ‘processor’, ‘memory’, ‘displaying’, ‘storing’, ‘receiving’ and so forth in claims as a way to recite an eligible machine or process.

The problem here is that over time all technology becomes generic – indeed it would be impossible to write patent claims if one could not refer to generic technology. If a ‘computer’ and ‘processor’ are deemed merely “generic technical language” and ‘storing’ and ‘receiving’ are “generic functional language” that cannot suffice to recite an eligible machine or process, then so too must ‘engine’, ‘resistor’, ‘heating’, ‘cooling’ and the like be insufficient, for these are all generic technical or functional terms.

The reason is that once a technology has become identified by a common noun, it is necessarily generic. An engine is generic with respect to the vastly different kinds of engines in use; heating is generic to myriad ways a substance can be heated. When a new invention is introduced it is considered a new distinct thing, but over time as it becomes common its name becomes generic. The escalator was once a specific (indeed trademarked) name, but eventually became a generic term for moving stairways. Almost every part of a computer that we consider generic today (eg, processor, touchscreen and memory) was at one time an entirely new invention and needed a new name. What is unique technology when first introduced becomes generic over time as competitors and others create similar products and an entire class of products is born and grows. In 1964 the computer mouse was invented, and there was only a single kind; today there are dozens. Thus, if the proposed framework is adopted as law it would mean that as any technological field matures, inventions actually become less patent eligible to the extent that they use the generic components in the field. This leads to the absurd result that an invention would be patent eligible when introduced early in the field, and then ineligible after the underlying technology becomes common and can be referenced using generic terms. Logically, a given invention cannot be eligible at one point in time and ineligible later on.

An escape route from Alice – from the inventor of the claims at issue

By Ian K Shepherd

As I read the Supreme Court’s Alice ruling on 19 June 2014 I recall thinking something like: “Beyond our case, this is not going to end well.” Now, in 2019, our challenge is to cease articulating the problem that is Alice and instead to focus on devising and promulgating a solution to the problem. I propose here an escape route from Alice based on first principles.

I propose that there be three tests of patent eligibility:

  • A test, already broadly accepted, that a claim needs to be “a useful discovery in the useful arts”, partially statutory, novel, inventive and sufficiently enabled.
  • A currently unspoken test that a claim cannot be unreasonably pre-emptive and/or a potential source of serious harm to the United States’ interests.
  • In respect of patent ineligibility summary judgment (PISJ) hearings, a sole test that the USPTO did not err in determining that a claim was/is inventive.

Test one

Arguably, only three aspects of this test require further comment. First, as is well understood by practitioners, the notion of “a useful discovery in the useful arts” is set out in Section 8(8) of the Constitution. Arguably, this phrase was used here to, in part, distinguish patentable inventions from “writings in science”. Second, “partially statutory” means the current Section 101 test but without the words “useful”, “new” and “any useful improvement thereof”. And third, specifically in respect of computer-implemented invention claims – the type of claims at issue in Alice – the inventiveness test element needs to include the question of whether the involved computer is intrinsic to the claim. This component is required to help ensure that the examiner of the claim at the USPTO is not unreasonably lenient in judging that the involved computer bestows inventiveness on a process or system that was, or could have been, previously performed with a pen and paper or in a person’s mind. The specifics of making this proposal intrinsic to the claim test is set out in the comments that Alice Corporation recently provided to the USPTO on its 2019 Revised Subject Matter Eligibility Guidance.

Test two

Until such time that it is explicitly recognised that a claim cannot be unreasonably pre-emptive in order to be patent eligible (see Gottschalk v Benson) and/or a potential source of serious harm to the United States’ interests, some patent ineligibility decisions by senior courts can be expected to have serious unintended consequences for the broader patent system. These conditions require new and creative thought in respect of issues such as:

  • which institution is best qualified to perform this assessment (an entity other than either the USPTO or a court?) and under what circumstances; and
  • whether the inventors/owners of patent claims judged to be patent ineligible on this test basis should be reasonably compensated for their economic loss.

Test 3

In PISJ hearings, as distinct from USPTO evaluations and court trials, not all then-relevant case facts are available for consideration. This is because such hearings primarily, if not exclusively, involve questions of law rather than fact. With USPTO examinations and court trials, all of the above-mentioned broadly accepted test conditions can be comprehensively assessed. Under the proposed regime, in respect of computer-implemented invention claims, PISJ hearings would be limited to resolving whether the involved computer is intrinsic to the claim in question.

While being beneficial in its own right, the proposed regime would achieve an even more substantial result – cessation of any future need for the Alice judicial exception tests for computer-implemented invention claims.

Ian K Shepherd is the founder and managing director of Alice Corporation, the inventor of the claims at issue in Alice and both defendant and litigant in Alice Corporation’s subsequent patent dispute with CLS Bank International

Alice and investment

Much of the argument about the impact of Alice is whether it has benefited or harmed investments in various technology sectors. Defenders of Alice argue that patents on software and financial innovations are unnecessary to spur innovation and that the overall reduction of patents in these industries – as well as the use of Alice as a tool to invalidate existing patents –helps to spur the creation of new companies and investments. Thus, these groups argue that no change in the law is necessary.

Critics of Alice take the opposing view, of course, and argue that the loss of patent protection has dramatically adversely affected research and investments – especially in the life sciences – and thus Congress should act sooner rather than later.

Yet for Congress to take up the daunting problem of amending Section 101, there needs to be more than arguments and prognostications by the stakeholders on opposite sides of the debate; there needs to be evidence of what impact, if any, Alice is having. That evidence is now available. Professor David O Taylor recently issued Patent Eligibility and Investment, which is the first research into the impact of the Supreme Court’s patent eligibility decisions (Bilski, Myriad, Mayo and Alice) on the investment activity of various types of investor. His findings generally support the need for Congressional action. Among the key findings are that:

  • In all, 74% of all investors in the survey agreed that patent eligibility is an important consideration in whether to invest in a particular company’s technology. This was true across all industries, from medical devices, biotech and pharmaceuticals (80% of investors responded that patent eligibility is important) to communications and software (73% of investors responded that patent eligibility is important). Those investors who were familiar with the Supreme Court’s patent eligibility decisions were statistically more likely to place greater emphasis on patent eligibility than those without such familiarity.
  • On average, investors would invest less in companies developing technologies for which patent protection was not available at all. The most affected industries were pharmaceuticals, biotech and medical devices, with approximately 77% of investors indicating that they would invest less in companies in these fields. The least affected were construction and software, with about 20% of investors indicating that they would invest less in construction, and about 35% indicating they would invest less in software and internet companies.
  • About one-third of the respondents were familiar with the Supreme Court’s eligibility decisions. Of these knowledgeable inventors, 49% reported that these decisions caused them to invest less, while 34% reported shifting investments away from pharmaceuticals, biotech, medical devices and software. Among knowledgeable investors, only 13% reported shifting their investments into software companies as a result of the court’s decisions.

These and other findings by Taylor support the view that the Supreme Court’s patent eligibility decisions have affected the investment landscape in a negative way, reducing investments across the board (at least among knowledgeable investors). While the court purports to merely state what the law is and to leave policy considerations to Congress, there can be no argument that its decisions are in fact policy setting. The Supreme Court – and the Federal Circuit below it – have decided that certain fields of innovations (eg, diagnostic methods, finance and gaming) are per se ineligible for patent protection. They have further decided that the ubiquity of computers, along with their general-purpose programmable functionality – the very thing that makes computers such powerful instruments – make inventions implemented by computers immediately suspect.

These positions combine to produce a perverse outcome: the more powerful our technologies become, the less likely such inventions are to be patent eligible. In the life sciences, the most powerful technologies replicate and control biological systems at molecular or lower levels – yet the court has held that such capabilities merely identify and apply laws of nature and natural correlations. In the computer sciences, we create ever more sophisticated algorithms (and hardware) that enable computers to make the same kinds of complex decisions as humans – but these inventions then are merely deemed to be abstract ideas because they appear to do what humans do mentally.

Given the constitutional mandate that the patent law is to promote innovation in all fields, not just some, it would seem that Congress should take seriously the prospect that it – not the Supreme Court – should set the technology policy for the United States.

Action plan

Few if any recent patent cases have had as great an impact as Alice Corp v CLS Bank International, sowing confusion and uncertainty for applicants across a range of sectors.

  • Alice has had a clear impact on the US patent system, with more than 960 patents invalidated in the courts and more than 60,000 abandoned applications since the ruling.
  • Section 101 has been used as a proxy for invaliding patents in court that would otherwise consider under obviousness or lacking in enablement or written description.
  • New Section 101 guidance represents a welcome change in tone from the USPTO but it is unclear whether it will survive scrutiny in court.
  • Recent new research has indicated that the decision has affected the investment landscape for IP-rich businesses in a negative way.
  • With patent owners still searching for certainty on what is patentable, legislative action may be the only long-term fix. Existing proposals offer a starting place but must be drafted to minimise the risk that the courts will apply existing interpretations of the judicial exceptions to the categories’ ineligible subject matter.

Unlock unlimited access to all IAM content