- I. INTRODUCTION
On October 30, 2008, the United States Court of Appeals for the Federal Circuit (CAFC or Federal Circuit) established a new standard for determining whether a process claim falls within the scope of the patentable subject matter. As a result, in contrast with its decision in State Street Bank, which drastically expanded the scope of patentable subject matter including business methods, the Bilski court significantly narrowed the scope of patentable subject matter.
Historically, in the United States, mathematical algorithms and business methods had not been subject to patent protection. However, unlike the patent law in Europe, United States patent law does not explicitly exclude such mathematical algorithms and business method claims from the scope of patentable subject matter. United States patent law requires an invention only to be novel, useful and non-obvious for patent protection. The State Street Bank court, therefore, relied heavily upon the language in the statute (35 U.S.C. § 101) and Congress’s intention, and then finally concluded that the essential characteristic of the subject matter is its practical utility. This logic seems to fit perfectly into United States patent law, but unfortunately, after State Street Bank, the United States Patent and Trademark Office (USPTO) has struggled with determining the quality of filed computer software and business method applications and has been criticized for issuing questionable computer software and business method patents.
From the international harmonization aspect, on the other side, Europe and Japan, which do not recognize business methods as such as a patent, have still kept a distance from the United States approach due mainly to the underlying policies, even though they have attempted to harmonize their approach to business method patents and computer software patents. Ironically, however, a decade after State Street Bank, the Federal Circuit seemed to be pressed to make a decision against narrowing the scope of the patentable subject matter in Bilski, and now, the new approach of the United States has seemed to compromise and come close to the European and Japanese approaches. The change in the U.S. approach might be largely affected by the high anxiety and criticism about the quality of previously issued computer software and business method patents. In other words, policy concerns may have caused this change. In his dissenting opinion in Bilski, however, Judge Rader warned that there is no legal basis for the new approach, and “[the Patent] Act has not empowered the courts to impose limitations on patent eligible subject matter beyond the broad and ordinary meaning of the terms” in § 101. Accordingly, U.S. courts still have unresolved matters, and they will struggle with the settlement of these matters.
In this article, we analyze the tests which have been adopted by the U.S. courts from legal and political perspectives in order to determine the scope of the patentable subject matter, especially for computer software and business method inventions. Even though the situations in Europe and Japan are quite different from the United States, they have struggled to delineate the scope of patentable subject matter as well, and have improved their approaches. Therefore, we also focus on these approaches so as to suggest an ideal approach and to promote international harmonization.
We start our analysis with the necessity of patent protection for computer software and business method inventions in Part II, and then review the state of patent law related to computer software and business methods in the Unites States, Europe and Japan, and the continuing issues confronting the respective Patent Offices in Part III. Part IV comparatively analyzes the three systems and considers the possible scope of the patentable subject matter with respect to the international harmonization of computer software and business method inventions. Finally, Part V discusses patentability, the next step for being granted a patent, of computer software and business methods in these countries.
- II. UNDERLING POLICY OF PATENT ELIGIBILITY
In the twentieth century, significant scientific and technological improvement have been observed, and computer-related technologies, including computer software and business methods, led rapid and drastic changes in modern society. In the meantime, the question of patent eligibility for computer-related inventions began to arise during the 1960’s and the early 1970’s. The question has been debated worldwide for almost 50 years. The major subject of debate has been whether computer software (programs) alone and business method inventions should be patentable subject matter. The main theory to deny the patent eligibility of computer software was that computer software was the same as a mental process or a step implemented in the head, it is merely a mathematical method. On the other hand, the theory supporting patent eligibility of computer software was based on the approach that although “a general-purpose digital computer may be regarded as but a storeroom of parts and/or electrical components[,] once a program has been introduced, the general-purpose digital computer becomes a special-purpose digital computer.” , There was another theory which took the middle ground of these theories: patent eligibility of computer software depended on the kind of software. For example, a computer program itself does not fall within the scope of patentable subject matter, but a computer program for automatically controlling generators would. After many debates and changes, the theory taking the middle ground seemed to become the mainstream theory in the world. In general, while Europe and Japan seem to have been reluctant to include computer software alone and business method inventions into patentable subject matter, the United States has shown tolerance toward such inventions, at least before Bilski. On the other hand, business methods were widely thought not to meet the patent eligibility until State Street Bank in 1998. After this case, things turned around in the United States, Europe and Japan, and then the door opened for business methods to be patented.
Thus, with the development of information technology (IT) and IT businesses, the threshold for patent eligibility seems to be getting lower, and both the computer software and business methods tend to fall into patentable subject matter in each country. Some commentators who oppose computer software patents point out that software’s novelty and utility generally reside in its mathematical algorithm, not in its physical structure. They argue that the algorithm cannot be patentable subject matter because it is a mathematical formula. A very short life time, which is a nature of a computer software invention, also supports opponents’ argument.
On the other side, others point out the necessity of permitting patentability because other forms of protection, such as copyright and trade secret, may be inadequate vehicles for the protection of computer software inventions. Patent protection is never destroyed by independent conception or reverse engineering of a claimed invention by a third party. Moreover, there is an opinion that the countries that have had the best patent systems have been the ones with the most dynamic economies. By providing an inventor the greatest patent protection, the United States has long been an enviable position in the world in this regard. It is apparent from the fact that the majority of the world’s great computer software-related companies, such as IBM, Microsoft, Apple and eBay, are American companies. Thus, since the patent system is basically intended to encourage improvement of industries and should be used to create attractive markets, excluding certain inventions from patentable subject matter merely because of computer software or business methods would cast doubt on the underlying principles of the patent system, especially in such a computer technology era.
Even if the inventions are categorized as a computer program and/or business method, the recognition that some of them still satisfy patentability requirements and are worth granting a limited-time monopoly has been clear for years. Although the market life of computer software inventions is very short, patent protection should be left as an option. Thus, the United States, Europe and Japan have recognized that computer software and business method inventions can be protected by patent, like other inventions. These attitudes are also consistent with Article 27(1) of the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPs), which states that all World Trade Organization (WTO) members must make patents "available for any inventions … in all fields of technology ….”
To take a radical position it is possible to establish a patent system in which the computer software and business methods inventions are regarded as already meeting patent eligibility and then are examined only for patentability. Now, however, at least the United States, the European Patent Convention and Japan do not adopt such a patent system. Why? Policy matters may perform a crucial function to lay the groundwork for the current patent system of each country. There are some kinds of inventions which have natures that are not suitable for the patent system at all, which grants a limited time and scope monopoly in exchange for disclosure of the inventions to the public. In essence, each patent system has some exceptions which are never protected by the system, such as abstract ideas and laws of nature, and these exceptions are virtually identical in each system even though it is slightly different in detail due to respective policy concerns. Thus, by setting forth the patent eligibility requirement, lawmakers represent that they do not condone a monopoly for certain types of inventions on the basis of the nature of the inventions, even though they satisfy the novelty and non-obviousness requirements.
Even though patent eligibility of the computer software and business method inventions is examined in the same way as other inventions, the appropriate test to decide whether a computer related-invention falls within the scope of patentable subject matter is another issue. In other words, to what extent is it proper to grant a time-limited monopoly in the case of the computer-related inventions? To delineate the proper scope, the United States, the European Patent Convention and Japan have struggled and eventually developed the respective tests explained in the following Part.
- III. PATENT ELIGIBILITY PRACTICES OF EACH COUNTRY
A. Patent Eligibility in the United States
The United States patent law is rooted in the patent clause of the Constitution and only 35 U.S.C. § 101 limits patentable subject matter as follows:
“[w]hoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent thereof, subject to the conditions and requirements of this title.”
As shown, § 101 just requires inventions to be “new and useful,” and does not substantially indicate what is out of the scope of the patentable subject matter. Therefore, interpretation of § 101 has been required to determine what falls inside and outside the scope. The Supreme Court in Chakrabarty held that “Congress intended statutory subject matter to ‘include anything under the sun that is made by man.’” , Historically, nonetheless, certain types of inventions have been excluded from patentable subject matter in the United States: laws of nature, natural phenomena and abstract ideas. These judicially created exceptions have been widely accepted. The underlying policy of these exceptions can be recognized in the words in Benson which addressed that “[p]henomena of nature, though just discovered, mental processes, and abstract intellectual concepts are not patentable, as they are the basic tools of scientific and technological work. … He who discovers a hitherto unknown phenomenon of nature has no claim to a monopoly of it which the law recognizes.”
Thus, in the United States, the fundamental principle of patent eligibility relied on the principle that basic tools of scientific and technological works, and abstract ideas are not suitable for a monopoly offered by the patent system. In this sense, the nature of computer software (programs) and business methods may be considered to be the basic tools of scientific and technological works and/or merely abstract ideas. Therefore, the view that such inventions should be treated the same as the judicially created three exceptions seems to have been predominant in the past.
Historically, the USPTO has denied patent applications for computer programs on the ground of nonstatutory subject matter, even though a program claim was drafted in the manner of a process, a machine or components. In such a situation, in 1966, the President’s Commission on the Patent System rejected the proposal that computer programs be patentable, and further suggested that patent protection for computer programs was not necessary, rather copyright protection was suitable. Relying on the Commission of 1966, the USPTO issued guidelines in 1968 (the 1968 guidelines) which stated that computer software, whether claimed as an apparatus or as a process, was unpatentable unless it was combined with a mechanical device. In short, the USPTO generally rejected the patent eligibility of computer software alone. The USPTO seemed to view computer software inventions, which contained or related to computer software, as unpatentable mental steps or unpatentable general purpose digital computers applying principles of mathematics, and then as not patent-eligible processes or machines. At the same time, the Court of Customs and Patent Appeals (the CCPA) took a different view from the USPTO. It considered that computer software inventions could be patentable subject matter. Thus, the USPTO’s view was not accepted by the CCPA. The Supreme Court in the 1970s, meanwhile, twice addressed whether inventions containing computer software were entitled to the protection of federal patent law. Both times, the Supreme Court sustained the USPTO’s rejections of the claims (under the 1968 guidelines) against reversals by the CCPA.
One of the two cases is Gottschalk v. Benson decided in 1972. The invention at issue was a process which converted a number expressed in binary coded decimal form into the same number expressed in pure binary form, for use in a general purpose digital computer. In this case, the Supreme Court considered whether an algorithm (an algorithm in computer programming) to convert binary coded decimal numbers into true binary numbers was a “process” within the meaning of the Patent Act § 100(b) and § 101. The court first clearly stated that natural phenomena, mental processes and abstract ideas are not patentable subject matter, “as they are the basic tools of scientific and technological work.” Then the Court held that the algorithm in question did not fall within the scope of patentable subject matter, because “[t]he mathematical formula … has no substantial practical application except in connection with a [general purpose] digital computer, which means that if the judgment below is affirmed, the patent would wholly pre-empt the mathematical formula and in practical effect would be a patent on the algorithm itself.”
The holding could be construed that computer programs are regarded as mathematical algorithms, and mathematical algorithms are considered to be abstract ideas, as a result computer programs (mathematical algorithms) are not patent eligible under §101. It seems to make the fourth category, computer programs (mathematical algorithms), as an exception of patentable subject matter. Another interpretation is also possible that a substantial practical application is required to be patent eligible. At least operating computer programs on a general purpose computer is insufficient to be regarded as a substantial practical application. Either way, it had been difficult for computer software inventions to become patent eligible, and thus, the Benson decision cast a shadow on such inventions.
Benson further established some important standards that affected subsequent cases. One of them was that, the Court simply introduced the requirement that “a process patent must either be tied to a particular machine or apparatus or must operate to change articles or materials to a ‘different state or things.’” Although the Benson Court never adopted that standard, it became crucially important in 2008. Another important standard was that judicially made exceptions, such as abstract ideas and laws of nature, would be equally applied to both a process claim and a product claim.
There might be doubts about the reasoning whether computer software inventions would ever become patentable subject matter, but the result seemed to be reasonable. In this case, as the majority noted, the claimed invention addressed only the simple mathematical procedures that could be performed without a computer, and claimed neither a field of use nor a practical application. Therefore, if Benson’s claimed invention was granted a patent, his exclusive right could prevail over uses of processes which he had not conceived, reduced to practice and described, indeed he had not invented. U.S. patent law never warrants such a broad scope monopoly.
The other case was Parker v. Flook in 1978, in which the Supreme Court again showed a negative attitude toward computer software inventions. The claim at issue described a method of updating an alarm limit which was a number. The method consisted of three steps; first, measuring the present value of the process variable (e.g., the temperature); second, using an algorithm to calculate an updated alarm-limit value; and third, adjusting the actual alarm limit to the updated value. The only difference between the prior art and the invention was the second step, the mathematical algorithm or formula. In this case, the Court confronted the issue whether a method for updating an alarm limit in a catalytic conversion process fell into the scope of the patentable subject matter.
The Court clarified that “a process is not unpatentable [subject matter] simply because it contains a law of nature or a mathematical algorithm,” but “the process itself, not merely the mathematical algorithm, must be new and useful.” Notably, the Supreme Court clearly denied the mathematical algorithms’ contribution to novelty, and treated such algorithms as if they had been a well-known part of the prior art. Further, the Court denied the identification of a limited category of use, and noted that post-solution application of a mathematical formula can not transform unstatutory subject matter into patentable subject matter, since it is against the spirit of patent law to easily avoid a § 101 rejection by just attaching some form of post-solution activity to a mathematical formula.
Then, although the Court seemed to have confused the patentability inquiry with the patentable subject matter inquiry, it held that the application at issue contained no patentable invention because once the claimed mathematical formula is assumed to be within the prior art, the claim contained no patentable invention. Thus, according to the holding, whenever a claim directed essentially to a mathematical formula, regardless if the formula is new or useful, it cannot be statutory subject matter, even though the solution is for a specific purpose or application. Since most computer software inventions are characteristic in computer program, algorithm and formula, the holding would be fatal for such inventions at that time. Thus, although the reasoning is based on novelty rather than patent eligibility, the Supreme Court expressed its intention that it has never allowed patent protection for computer software inventions.
On the other hand, the Court gave a favorable review on a claim containing mathematical algorithms denying the view that merely containing such an algorithm causes unpatentability. Instead, the Court found that “a patent claim must be considered as a whole.” The “as a whole” analysis has been taken over by the following cases in the United States, and it has been seen in the European and Japanese approaches.
At that time of Benson and Flook, the United States government viewed a patent as a tool relied upon by big companies to squelch competition. Therefore, government policy emphasized enforcing the antitrust laws and establishing patent antitrust policies rather than encouraging innovative activities. Thus, the Flook Court’s decision resulted in being strongly influenced by the government policy at that time.
The negative attitude of the Supreme Court for computer software inventions continued until the Diamond v. Diehr case was issued in 1981. The year before Diehr, however, there was an indication of the imminent turning point. In 1980, the Supreme Court published one significant 5-4 decision. In Diamond v. Chakrabarty, the Court found that, considering the committee reports accompanying the 1952 Patent Act, “anything under the sun” was included in statutory subject matter, and held that “Congress plainly contemplated that the patent laws would be given wide scope.” The Court’s recognition of such broad scope of § 101 might have largely influenced Diehr, and triggered a change in its earlier denial of patent eligibility of computer software inventions. Further, a political change would affect the decision in Diehr: the so-called pro-patent era had begun with the election of Ronald Reagan as President of the United States. The Reagan Administration ended the era of antitrust ascendancy and realized that innovation was one of the most important American “products.” Thus, a strong patent system became an essential factor of Reagan's policy. To enhance the patent system and promote national uniformity in patent law, the Reagan Administration quickly started the work to establish the CAFC as a national court of appeals for all patent matters.
Under these circumstances, in Diehr, unlike Benson and Flook, the Supreme Court affirmed the decision of the CCPA, which reversed the USPTO’s rejection that relied on Benson, in a 5 to 4 ruling. In this case, the question was the patent eligibility of a claimed method and programmed digital computer for determining how rubber should be heated in order to be best cured by using a computer program employing the Arrhenius equation. Contrary to Flook, a mathematical formula (the Arrhenius equation) had long been used to calculate the cure time, and the process of constantly measuring the actual temperature inside the mold was indeed a characteristic feature. By citing Benson, the Supreme Court ruled that “transformation and reduction of an article ‘to a different state or thing’ is the clue to the patentability of a process claim that does not include particular machines.” Then the Court found that the claims involving the transformation of an article (raw, uncured rubber) into the different stage or things, and the industrial processes such as curing rubber at issue were the types which had historically been eligible to be protected by the patent laws. Moreover, the Court differentiated this case from other computer-related inventions, such as Benson and Flook, by noting that Diehr sought to patent an industrial process for the molding of rubber with transformation of rubber to a different state or thing, it had not attempted to patent an unpatentable law of nature or mathematical algorithm, such as Benson and Flook.
One of the significant outcomes in Diehr would be that a certain type of test, the so-called transformation test in Bilski, was established to determine whether the claimed process, without reciting a particular machine, falls within the scope of the patentable subject matter. Remarkably, to quell preemption, the majority looked for a clue in the effects upon physical objects. The other significant outcome was an indication that the patent eligibility of the claimed invention never should be denied due simply to containing a mathematical algorithm. Thus, backed by pro-patent policy, Diehr seemed to shake off a feeling of anxiety that an invention including a computer program would never be patent eligible.
The Court, however, left some questions unanswered. For example, the majority did not explain in detail what the meaning and scope of the term “transformation.” Moreover, as shown in Justice Stevens’ dissenting opinion, the majority decisions in Flook and Diehr appear to contradict each other in fundamental ways. In other words, there seems to be no reasonable logic to explain why although the method of updating the curing time calculation was strikingly reminiscent of the method of updating alarm limits that Flook sought to patent, the court reached opposite results. In trying to make consistent with Flook, a claim directed essentially to a mathematical formula, even if it was new and useful, seems still outside the statutory subject matter. Moreover, since this case focused only on a process claim without reciting a particular machine, and the meaning of a “particular” machine seems to have been unclear, the requirements for patent eligibility of an apparatus claim or a process claim including a general purpose machine were left unresolved.
From Benson to just after Diehr, the CCPA created and further modified its own test in order to remain consistent with the Supreme Court decisions. After Benson, in In re Freeman, to decide whether a computer software invention is inside of the patentable subject matter, the CCPA established a two-prong test: “[f]irst, it must be determined whether the claim directly or indirectly recites an ‘algorithm’ in the Benson sense of that term,” and “[s]econd, the claim must be further analyzed to ascertain whether in its entirety it wholly preempts that algorithm.” In In re Walter, the CCPA added the term “as a whole” to the second inquiry after the Flook decision.
After Diehr, it refined and finalized the two-prong test in In re Abele. With the Abele decision, the CCPA completed its cumbersome Freeman-Walter-Abele test for judging patent eligibility of computer software inventions. After Diehr, since the USPTO would not take the case containing a computer software invention to the Supreme Court anymore, the CCPA, the CAFC (a descendant of the CCPA) and the USPTO basically applied the Freeman-Walter-Abele test to determination the patent eligibility of a computer software invention until the CAFC’s decision in In re Alappat in 1994, which finally discarded the test. In the meantime, however, while generally the CCPA and the CAFC were in favor of a computer software invention, the USPTO firmly maintained a “strong institutional bias” against patenting computer software inventions broadly, despite the Reagan Administration’s pro-patent position and Diehr.
In 1994, the CAFC finally settled the dispute about patent eligibility of computer software (algorithms) between the USPTO and the appellate court in In re Alappat, en banc. The apparatus claim being described in a means-plus-function claim form related to a mechanism for producing smooth waveform display on an oscilloscope (OSC) screen. By scrutinizing precedents, Diehr, Flook, and Benson, the CAFC noted that “the Supreme Court never intended to create an overly broad, fourth category of subject matter excluded from § 101” other than laws of nature, natural phenomena, and abstract ideas, rather the Court attempted to exclude “certain types of mathematical subject matter, [which] standing alone, represent nothing more than abstract ideas until reduced to some type of practical application, and thus that subject matter is not, in and of itself, entitled to patent protection.” Further, the CAFC found that a “general purpose computer in effect become[s] a special purpose computer once it is programmed to perform particular functions pursuant to instructions from program software.” Then the CAFC held that the proper inquiry is “whether the claimed subject matter as a whole is a disembodied mathematical concept, which in essence represents nothing more than a ‘law of nature,’ ‘natural phenomenon,’ or ‘abstract idea.’” In sum, the new test chosen by the CAFC shows that a computer software invention did not need special analysis regarding the patentable subject matter anymore. By applying the new approach, the court concluded that because the claim as a whole was directed to some kind of combination to a machine, ‘this is not a disembodied mathematical concept which may be characterized as “an abstract idea,” but a specific machine to produce a useful, concrete, and tangible result.’
As for the Freeman-Walter-Abele test, the Alappat court virtually discarded the test. Since the claimed patentable subject matter should be considered as a whole, the Alappat court emphasized that it is unnecessary to focus on each claimed element and inquire whether a claim contains any mathematical algorithms, which standing alone would not be patent eligible. In addition, as noted by Judge Rader in his concurring opinion in Arrhythmia, the first inquiry in the Freeman-Walter-Abele test had been criticized, partly because the term “mathematical algorithm” remained vague.
Thus, the majority in Alappat completely dispelled any fear about the computer program exception. In addition, the court seems to have begun to weigh features of a computer program (software) rather than the particularity of hardware by implying that a general purpose computer could turn out to be a special purpose computer. The fact that inventions rely heavily on mathematical algorithms did not confer an adverse inference in assessing their suitability for patent protection. On the other hand, whether the term, “a useful, concrete, and tangible result,” is used as expanding the meaning of “practical application” in Benson and “transformation of an article into a different stage” in Diehr, or just paraphrasing these words has been uncertain. This phase, however, pulled the trigger of significant expansion of patentable subject matter later on.
During the same period, further dramatic changes which have had a great influence on the computer software industry, were brought by the following CAFC’s decisions: In re Scharader, In re Warmerdam, In re Lowry and In re Beauregard. The indication to focus more on computer software in Alappat became clear in these following cases. In Scharader, finding that the Hotel Security Checking case, which was often cited as establishing the business method exception to patentable subject matter, actually discussed the obviousness of the system at issue, the dissent from Judge Newman argued that there was no legal basis for the “business methods exception” to patent eligibility. In the middle two cases, Warmerdam and Lowry, the Federal Circuit found that computer programs or data structure existing in a memory could be statutory subject matter, since data stored on a disk or memory provided a physical change to the disk or memory. As a result, these decisions that computer programs embodied in a tangible medium are entitled to be patentable subject matter forced the USPTO to abandon its “printed matter” practice, on which it had relied to deny the patent eligibility of a computer-readable media claim. The last case, Beauregard, was a decisive case to discard the “printed matter” doctrine. Consequently, hardware and software companies like IBM could argue direct infringement of their software products per se without the help of contributory infringement analysis. In other words, emerging software business exploded at that time, and the CAFC and the USPTO were forced to change their attitudes toward computer software itself.
Following these decisions by the CAFC, the USPTO ended up issuing proposed Guidelines for Examination of Computer Implemented Inventions (1996) which was eventually incorporated in the Manual of Patent Examining Procedure (MPEP) as section 2106 in 1998. Under this Guideline, computer-implemented inventions were generally regarded as patentable subject matter as far as they had a “practical application.” Furthermore, the Guideline promulgated that although per se computer programs were not patentable subject matter, computer programs stored on a computer-readable medium could be. Nonfunctional descriptive material (e.g. abstract ideas), however, could not be statutory subject matter even stored in a computer-readable medium. At the same time, the USPTO was forced to show tolerance toward “methods of doing business,” which had been categorized as nonstatutory subject matter in the MPEP, because of the emergence of patent applications on e-commerce. As a result, the USPTO deleted the business method from the nonstatutory subject matter in MPEP (1996). Thus, the USPTO decided that methods for doing business “should be treated like any other process claims.”
In the meantime, the CAFC had a chance to address the so-called “mathematical algorithm” and “business method” exceptions to statutory subject matter in State Street Bank. In this case, Signature Financial (Signature) who was a patent holder of a “Hub and Spoke” method of running mutual funds appealed a decision from the district court, which granted State Street Bank's motion for summary judgment and found Signature’s apparatus patent (described in means-plus-function form) invalid due to the lack of patent eligibility. In reversing the decision of the district court, the Federal Circuit revisited Arrhythmia and Alappat, and determined that although an invention which consisted solely of a mathematical algorithm that represented nothing more than an abstract idea was not patentable, a mathematical algorithm could be patent eligible as long as it was applied in a “useful” way. Emphasizing that software or other processes that yield a useful, concrete and tangible result should be considered eligible, the court held that the patent in question fell within the scope of patentable subject matter because it produces “a useful, concrete and tangible result.” By focusing heavily on the term “useful” in § 101 rather than precedents that had tried to avoid preemption due partly to computer software’s abstractiveness, the scope of the statutory subject matter was significantly enlarged by the Federal Circuit.
As for the business method exception, the court explicitly wiped out this “ill-conceived exception,” and concluded that there is “no longer applicable legal principle” and “business methods have been, and should have been, subject to the same legal requirements for patentability as applied to any other process or method” since the 1952 Patent Act.
Moreover, the subsequent decision in AT&T advanced the trend of expanding the scope of statutory subject matter. In this case, the Federal Circuit explicitly rejected Excel’s argument based on Diehr that a physical transformation is required for patent eligibility by reasoning that physical transformation “is not an invariable requirement, but merely one example of how a mathematical algorithm may bring about a useful application.”
After the State Street Bank and AT&T decisions, the USPTO was forced to revise the MPEP again in order to remain consist with these decisions. As a result, there is little doubt that computer software and data structures can be considered patentable subject matter in the United States as far as they produce useful, concrete and tangible result to have a practical application. Whether the computer software relates to methods for doing business is an irrelevant inquiry.
The logic noted by State Street Bank and AT&T relied upon the terms in the Constitution and U.S. patent law and was very natural. On the other hand, the Federal Circuit and the USPTO seem to have virtually disregarded Supreme Court precedent, which asked for a connection between software and physical objects so as to avoid preemption. However, the existence of the judicially created exceptions should not be ignored when patent eligibility is discussed, because these exceptions are rooted in the spirit of the patent system; certain inventions are unsuitable for patent protection, no matter how useful they are. Nonetheless, the Supreme Court denied certiorari twice, in the State Street Bank and AT&T cases, which suggests that the Court had tolerance and deferred the role of interpreting the scope of § 101 to the Federal Circuit. A series of these movements show that the whole country, including Congress and the courts were strongly promoting the pro-patent policy at that time.
The situation, however, changed completely. After the State Street Bank decision, the expansion of patent protection for business methods and computer software had become accepted practice in the United States, however, questionable business method patents had been issued and the negative impact of patent trolls emerged. In the meantime, the Supreme Court began to express its discomfort with business method patents. In 2006, in eBay, Justice Kennedy noted the “potential vagueness” and “suspect validity” of some business method patents in his concurring opinion. Further, in Laboratory Corporation of America Holdings v. Metabolite Laboratories, Inc., the Supreme Court briefly commented on the issue of patent eligibility as if as a warning. Referring back to the old cases, such as Cochrane and Benson, in his dissenting opinion, Justice Breyer strongly criticized the CAFC's recent decisions as possibly being inconsistent with Supreme Court precedent. He clearly stated that the Supreme Court has never said that a process is patentable if it produces a “useful, concrete and tangible result” in State Street Bank.
On September 20, 2007, the year after Metabolite, the CAFC released two opinions pertaining to the patent eligibility of computer software inventions, which somewhat differ from those before them. In the first case, In re Comiskey, the application claims a method and system of arbitration involving legal documents, such as wills and contracts, which does not necessarily require the use of any technology, and the court held that a “method and system for mandatory arbitration involving legal documents” was classified as an unpatentable subject matter because it was a “mental process.” To reach this conclusion, the court reconfirmed the fundamental principle that an abstract idea itself would never be patentable subject matter, and noted that an abstract idea employed in a process can be patent eligible when it has a practical application, and further when the process was tied to a machine or was transforming materials to a different stage. Thus, contrary to State Street Bank and AT&T, the court expressed its intention to require some sort of physicality by relying heavily on Benson, Flook and Diehr.
In the other case, In re Nuijten, the CAFC considered whether a signal is patentable subject matter. Notably, the court ruled that “[i]f a claim covers material not found in any of the four statutory categories, that claim falls outside the plainly expressed scope of § 101 even if the subject matter is otherwise new and useful.” Then, since a “signal” would not fit any of the four § 101 categories, the court held that claims seeking to patent any “signal” were rejected. Patent eligibility of a signal has become dramatically important because software is being developed using less reliance on physical entities. Therefore, this decision would have important implications for computer software inventions, especially Internet software.
Thus, Justice Breyer’s dissent in Metabolite seems to have had a great impact on the CAFC Judges. These cases are also particularly remarkable because the “mental steps” doctrine, which rejected patent protection for any process that could be performed in the mind, was developed in the United States during the 1970's and has since been discarded, and because the § 101 categories were always said to require the broadest possible interpretation. Therefore, conformity with precedents would cast a shadow of doubt on this reasoning, but it is at least clear that the CAFC began to change its policy against the patentable subject matter of computer software inventions.
Almost one year later, on October 30, 2008, the CAFC (en banc) finally made clear the drastic change in policy by publishing its latest decision related to a “process” claim for computer software inventions, In re Bilski. The inventions at issue were a method for hedging risk in the field of commodities trading by entering into contracts to purchase and sell commodities at fixed prices. In that case, the majority stated that the useful, concrete, and tangible result test, which was first set forth in Alappat and followed by State Street Bank, is inadequate and insufficient to determine patentable subject matter because of lack of intention to supplant the Supreme Court’s test.
Instead, the majority set forth the following test: a claimed process is patent eligible subject matter if “(1) it is tied to a particular machine or apparatus, or (2) it transforms a particular article into a different state or thing.” Regarding to the meaning of an “article,” the court further noted with reference to precedents that any physical object or substance, or an electronic signal representative of any physical object or substance would constitute an “article.” In addition, the court provided some guidance. First, by citing Benson, “the use of a specific machine or transformation of an article must impose meaningful limits on the claim’s scope to impart patent-eligibility.” Second, by citing Flook, “the involvement of the machine or transformation in the claimed process must not merely be insignificant extra-solution activity,” such as data gathering or outputting, and transformation must be central to the purpose of the claimed process. As mentioned by the majority, although a process satisfying the machine-or-transformation test will surely satisfy the useful, concrete, and tangible result test, the useful, concrete, and tangible result test is still inadequate and insufficient. In short, the useful, concrete, and tangible result test, relied heavily on by the USPTO after State Street Bank and AT&T, could regard inventions as patent eligible which are not patent eligible under the machine-or-transformation test. Thus this test substantially narrows the scope of patent eligibility of computer software inventions including business methods, and thus it seems to avoid further criticism from the Supreme Court and industry. As a result, many software implemented process patent claims could be invalid. A question left to be determined is what effect this would have on “apparatus” claims for computer-implemented software.
Now, after Bilski, are there any changes in the USPTO and the federal district courts? On January 7, 2009, the Deputy Commissioner of Patent Examination Policy of the USPTO issued a memorandum stating that the guidelines for examination will be redrafted to reflect the machine-or-transformation test. Until the new guideline is issued, the memorandum announced that examiners who determine the subject matter eligibility of process claims should follow the current guidelines appearing in MPEP 2106 with modification based on the two-prong test in Bilski. As a result, examination in the USPTO has substantially been done under the machine-or-transformation test.
The Bilski decision also has incredible impact on the Board of Patent Appeals and Interferences (BPAI) of the USPTO. For example, from Bilski to the end of March, 2009, the BPAI has issued 17 decisions that reference Bilski and patentable subject matter under § 101, and in 16 cases, the BPAI either affirmed an examiner's § 101 rejection (eight cases) or entered a new ground of rejection under § 101 (eight cases), raising the § 101 issue sua sponte. In the remaining case, the BPAI remanded asking the examiner to consider whether the claims were patent eligible under § 101. Thus, the number of cases that were rejected on the ground of § 101 is dramatically exploding. Notably, in Ex Parte Cornea-Hasegan (on January 13, 2009) and Ex Parte Atkins (on January 31, 2009), the Board applied the machine-or-transformation test not only to process claims but also to non-process claims, and then rejected the patent eligibility of a “computer readable media” and a “system for converting the domain name.”
Significantly, the board in Ex Parte Cornea-Hasegan reasoned by citing Diehr that “[a]lthough the claim recites a ‘computer readable media’ and ‘hardware,’ those elements were seen as insignificant extra-solution activities” and they do “not add any practical limitation to the scope of the claim,” as a result, “[s]uch a field-of-use limitation is insufficient to render an otherwise ineligible claim patent eligible.” Following Lowry, since the Board pointed out that in the claimed invention, “the data represents information about an abstract floating-point number,” a computer readable media storing functional computer programs might reach a different result. For example, there is a Board decision stating that the computer-readable media claim, which represents a number of software components (modules), was statutory under Lowry (without citing the Bilski test). To be regarded functional (statutory), therefore, adding some structures may be required even in a computer-readable medium claim.
Moreover a computer-readable medium claim is now in face of new danger. Relying on Nuijten, such claims began to be rejected on the grounds of encompassing non-statutory subject matter, a “signal.” Whether a claimed invention covers a signal is determined based on not only claims, but also the Applicants’ disclosure and Specifications. In this Internet era, computer software is reasonably intended to be conveyed by signals (carrier waves) via the Internet. Therefore, most specifications would disclose the program might be embodied in a tangible medium (e.g. a CD-ROM) and intangible medium (e.g. a carrier wave and signal). Recent Board decisions seem to argue that these Internet technologies are not suitable for patent protection at all, even though they are really technical in nature. If so, as discussed in Part IV, this might be completely opposite to the public interest, international harmonization and TRIPs Article 27. Did the Bilski and Nuijten courts expect such consequences?
It is also noteworthy that, relying on the notion in Bilski that “a claimed process wherein all of the process steps may be performed entirely in the human mind … would not be patent-eligible under § 101,” a couple of Board decisions rejected applications under the mental process doctrine. Thus, recent BPAI decisions drastically narrow the scope of statutory subject matter so that most computer-readable mediums and mental processes would be effectively excluded from statutory subject matter.
On the other hand, one federal district court also published a decision on March 26, 2009 which conformed to Bilski. In Cybersource Co. v. Retail (N.D. Cal), the question was whether the patented invention, method and system claims for detecting fraud in a credit card transaction between a consumer and a merchant over the Internet, was within the scope of patentable subject matter under § 101. In applying the machine-or-transformation test, one of the issues was “whether recitation of ‘over the Internet’ suffices to tie a process claim to a particular machine.” Judge Patel has enumerated at least three reasons why reciting “over the Internet” is insufficient: preemption, insignificant extra-solution, and not imposing meaningful limits on the scope of claimed invention. Finally, she granted summary judgment in favor of defendant due to invalidity under § 101. Therefore, adding the term “Internet” to a claim so as to import patent eligibility into the claim would be doomed to failure. The other question was whether the “Beauregard” claim is a product claim to which the machine-or-transformation test does not apply. Judge Patel disagreed with the applicant’s argument, and noted that even if the patent contained Beauregard claims, there is “no legal doctrine creating a special ‘Beauregard claim’” exception in applying the test. Interestingly, after rejecting the argument, she delivered some observations on business method patents after Bilski. She noted that the business method patents “swelled” in number following the State Street Bank decision, but Bilski “suggests a perilous future for most business method patents.”
As if supporting Judge Patel’s comment, just before Cybersource, two cases regarding business method patents were released by the CAFC and a California federal district court respectively. One concerned methods for performing tax-deferred real estate exchanges, and the invention was granted under the useful, concrete and tangible result test (Fort Properties). The district court, however, ruled that the present patent was non-statutory subject matter under the Bilski test. The other was directed to a method and paradigm for marketing, selling and supporting special kinds of products (In re Ferguson). The claimed method was rejected by the Board because it was directed to an abstract idea. On appeal, the Federal Circuit held that the invention was patent eligible under the Bilski test. In these two cases, since the claimed inventions were method claims, the transformation test was a major topic of discussion. Both courts expressed similar reasoning: creating deedshares, and organizing business or legal relationship in the marketing company cannot be considered an article in the meaning of the transformation test. Thus, such pure business factors, which might be useful and statutory under State Street Bank, are difficult to be found as an “article” under Bilski, since, as Bilski stated, “they are not physical objects or substances, and they are not representative of physical objects or substances.” Although the Federal Circuit denied Applicant’s argument that “the method claims are directed to business methods and should be treated as statutory subject matter in light of State Street” because the subject matter in State Street Bank was not a method, but a machine, there is a question about how much the claim category means to the Bilski test. The machine for business purpose would still be outside the subject matter, unless it recites a specific machine that imposes meaningful limits on the claim scope. Consequently, since most business method inventions are characterized by a method for doing business (not a machine), as Judge Patel stated, Bilski seems to revive the groundless business method exception from the past.
The district courts’ and BPAI’s focus on Bilski have been primarily in computer software applications, but there are many that see a potential for Bilski to be broadly applied to other areas of technology. In reality, the Federal Circuit itself began to adopt the machine-or-transformation test not only to process claims of computer software but also to process claims of diagnosis. On December 19, 2008, the court issued a one-paragraph decision in Classen Immunotherapies v. Biogen. The court affirmed the district court's grant of summary judgment which noted that claims at issue are invalid under § 101 by reasoning that ‘[i]n light of our decision in In re Bilski, … Dr. Classen's claims are neither “tied to a particular machine or apparatus” nor do they “transform  a particular article into a different state or thing.”’
After this decision, on January 28, 2009, Mr. Bilski petitioned the Supreme Court for a writ of certiorari asking it to determine whether the machine-or-transformation test for patentable subject matter is the correct test or not, and on June 1, 2009, the Supreme Court granted the petition. Now Medistem has submitted its Amicus brief supporting certiorari in Bilski. In the brief, Medistem argues that “[t]he Federal Circuit’s limiting the scope of patentable subject matter for ‘process’ inventions in Bilski casts a cloud of uncertainty as to whether Medistem and other biotech companies can continue to protect with patents their inventions relating to methods of diagnosing causes of diseases and methods of selecting beneficial treatment protocols.” Medistem argues that “biotech companies lose the ability to obtain enforceable patent protection on diagnostic and treatment methods, and the ability to attract investment capital will sharply decline.” Medistem, however, is not the first to address this kind of fear. In his dissent from the en banc Bilski opinion, Judge Rader cogently noted that denying patent protection for methods of using discovered biological or physiological correlations will “undermine and discourage future research for diagnostic tools.”
Thus, in the United States, the scope of the patentable subject matter toward computer software inventions has drastically altered based mainly on the development of computer technology and changes in the administration’s patent policy. On the other hand, the Supreme Court’s current decisions addressing patent matters, such as eBay, MedImmune, KSR and Quanta, may signal that it will continue to adopt so-called pro-patent policy. In these situations, Bilski restricted the scope of patentable subject matter so narrow that it is similar to the 1970s or 1980s. Whether the machine-or-transformation test set forth in Bilski can be the test that underpins the competitiveness of the computer software industry should be examined.
B. Patent Eligibility in Europe
1. The European Patent Convention
In contrast to the United States, in Europe, there are statutory limitations for patentable subject matter. Even though there is no direct definition for “invention” or “patentable subject matter,” under the European Patent Convention (EPC), Article 52 governs patentability and patent eligibility as follows:
(1) European patents shall be granted for any inventions which are susceptible of industrial application, which are new and which involve an inventive step.
(2) The following in particular shall not be regarded as inventions within the meaning of paragraph 1:
(a) discoveries, scientific theories and mathematical methods;
(b) aesthetic creations;
(c) schemes, rules and methods for performing mental acts, playing games or doing business, and programs for computers;
(d) presentations of information.
(3) The provisions of paragraph 2 shall exclude patentability of the subject-matter or activities referred to in that provision only to the extent to which a European patent application or European patent relates to such subject-matter or activities as such.
Apparently, the methods for doing business and programs for computers seem to be absolutely barred from patentable subject matter under Article 52(2) of the EPC. Although Article 52(3) does not explain what is meant by “as such,” the term “as such” have been used as magic words that have enabled the European Patent Office (EPO) to grant patents for business methods and computer program inventions. The guidelines for Examination in the EPO (the EPO guidelines) actually state that, under Article 52(3), the exceptions listed in Article 52(2) have to be interpreted narrowly, and thus the subject-matter or activities on the list are patentable unless the European patent application or patent relates to them as such. Therefore, an invention having both a concrete and a technical character, which are or may be implemented by computer programs is or may be an “invention” in the meaning of Article 52(1).
In Europe, a “computer-implemented invention” has been used instead of a software patent in order to avoid misleading applicants to assume that source/object code is patentable. The term computer-implemented inventions indicates “claims which involve computers, computer networks or other conventional programmable apparatus whereby prima facie the novel features of the claimed invention are realized by means of a program or programs.” Such claims, for example, may “take the form of method of operating said apparatus, the apparatus set up to execute the method, or … the program itself.”
a. History of the Article 52(2) and (3) EPC
Unlike the patent laws in the United States and Japan, in the EPC, programs for computers and methods for doing business are explicitly drafted in the exclusion list. The exclusion of computer programs arose during a period of significant legal instability, toward patent system integration.
Prior to 1973
Prior to the 1960s, each nation of Europe had developed distinct legal frameworks for the respective patent systems, which reflected historical backgrounds and philosophies relating to “industrial property.” The key nations to the European patent system are the United Kingdom (the U.K.), France and Germany. In the United Kingdom, the patent law was determined by the constitutional history of that nation. In the case of France, it had basically depended on the patent law developed in 1790, which was a fundamental of French patent law until the late 1960s. Germany lastly developed its patent system in the nineteenth century with the foundation of that nation-state. Other European nations tended to follow one of the three established models, the U.K., French or German, to varying degrees.
At this period, the substantive feature common to all European patent laws was an emphasis on novelty and invention; patents should only be granted for new inventions, therefore the differences in the nations was provided by the additional factors required for patent eligibility. In the U.K., due to the historical background that a “manner of new manufacture” was defined as an invention, an invention must result in the production of some “vendible product.” In addition, utility (working the way it was supposed to and providing some useful result) was required to grant a patent. Thereafter, these requirements were known as the novelty, utility and inventive step approach.
On the other hand, in Germany, courts construed the concept of patentable invention to require a technical effect bringing an advance in art. Therefore, “a technical progress/advance” was required to be patent eligible, apart from the requirement of inventive height (the degree of invention). Thus Germany took a different approach from the U.K. to assess patent eligibility. As a result, in Europe, these two approaches (the British approach and the German approach) were dominant until the European Patent Convention was issued. At that time, however, a computer program and a method for doing business had not been recognized as exclusions from the statutory subject matter.
In the 1960s, Europe entered a period of integration for harmonization of patent systems. In the meantime, a draft Patent Cooperation Treaty (PCT) was published in 1967, and France enacted its revised patent law in 1968. Notably, the new French Patent Law clearly denied the patent eligibility of “programs of a calculation machine.” Although the new law remained silent on patent eligibility of a computer program itself, the basic direction of the new law seemed to be similar to that of Benson. The movement for establishing the European Patent did not stop the whole time, and after all sorts of troubles, difficulties and changes, a European System for the Grant of Patents (1970 Draft) was approved by the Luxembourg Inter-Governmental Conference. Basically, the British approach was incorporated in the 1970 draft, as shown in Article 9 of the 1970 Draft relating to patentable subject matter.
Article 9(1) was identical to Article 52(1) of the current EPC. Article 9(2), corresponding to Article 52(2) of the current EPC, was influenced by the PCT and as a result listed items that were explicitly excluded from the inventions under Article 9(1), even though the “programs for computers” and “methods for doing business” were not listed at that time. In the case of the PCT, the reason for the exclusion was based on a relatively practical matter, that is to release international searching authorities from having to undertake a prior art search in the case of some particularly difficult items. The underlying concept of the exclusion was not directly related to the philosophy, purpose or justice of the patent system, but importing the exclusion to the 1970 Draft might somehow indicate that the Convention tried to articulate the scope of the patentable subject matter with the negative list. Thereafter, the list was enlarged in 1971 to include programs for computers, and in 1973, just before adoption of the final EPC, Germany proposed adding a third article to provide an interpretive rule for the excluded list: only where an invention falling into the excluded items as such would be denied as not being an invention in the meaning of the EPC. Finally, On October 5, 1973, the Convention on the Grant of European Patents (EPC) was signed in Munich and became effective in member states in 1977. Article 9 of the 1970 Draft was codified as the current Article 52 of the EPC.
The EPC was finally signed in 1973, but before and even after that, there was controversy over inserting “programs for computers” on the lists which excluded subject from patent eligibility under Article 52. On the other hand, a debate on “methods for doing business” seems to not have been paid much attention. According to the General Rapporteur who prepared the final EPC draft, the reason for that exclusion was to avoid a particularly controversial issue, but whether the intention was merely to prevent the issue from digressing from the EPC or to settle the issue later remains unclear. In reality, when the time to make a decision regarding whether to include computer programs in the EPC's exclusion clause arose, member states had not yet shown a clear position on whether computer programs were patent eligible. Even France had not explicitly excluded a computer program as such under its Patent Law of 1968.
In 1973, just before the signing of the EPC, two remarkable decisions were published: one from the Cour d'Appel de Paris, the other from the Bundespatentgericht (German Federal Patent Court). In May of 1973, the Cour d'Appel de Paris found that programs of a calculating machine contained computer programs, then held that the legislature had intended to exclude such programs from patentable subject matter under the Patent Act of 1968. The Court reasoned that under the Act, rejection for patent ineligibility depended upon whether an invention was “manifestly” lacking in technical character; the legislature’s explicit exclusion of computer programs from patent eligibility suggested that the legislature had regarded such programs as clearly lacking technical character. Thus, from the Court’s view, the legislature concluded that all computer-implemented inventions lacked technical character, and consequently that they were completely excluded from patentable subject matter. Even though the reasoning is different, the holding that an invention containing computer programs is outside the statutory subject matter might be analogous to that of Benson in 1972.
On the other hand, soon after the decision of the Cour d'Appel de Paris, the German Federal Patent Court, the Bundespatentgericht, issued an opposite opinion regarding the patent eligibility of computer programs. The invention at issue was a method of converting decimal numerals into pure binary numerals, nearly identical to Benson’s invention. The German Federal Patent Court relied on the German Federal Supreme Court's holding in the Rote Taube (Red Dove) case, which indicated that there was no need for technicality for patent eligibility in the case of a biological invention, and held that because the invention constituted an application of technical means (parts of a data processor) in order to achieve a technical purpose (conversion of decimal to binary numbers), it possessed technical character, and thus was patent eligible. Since the holding seems to weigh on the result, such as usefulness, its approach may be comparable to the decision in Alappat and State Street Bank. Thus, although the German and French Courts both applied a technical character approach, they reached opposite conclusions.
Even applying the same approach, technical character, the completely opposite results shed light on the fact that each nation had a different perspective on patent protection for computer-implemented inventions. In contrast, the U.K. was the only nation that adhered to its original approach, a “manner of new manufactures” approach. At a minimum, there was no unified opinion or approach to assess the patent eligibility of computer-implemented inventions until the EPC was signed in 1973.
Just before the enactment of the EPC in 1977 and prior to other member states, Germany and the U.K. fundamentally revised their patent laws in order to conform to the EPC, and as a result, both their patent laws had provisions substantively the same to Article 52 EPC. In this regard, Germany accepted the U.K.’s inventive step approach and discarded its own technical character approach. However, even though the revisions were implemented, the patent eligibility of computer-implemented inventions was still problematic. In the meantime, in 1984, the EPO established a working group to study the controversial problem, and the next year, the Guidelines for Examination for the EPO were revised to incorporate the working group’s suggestion that the EPO should consider technical character as one of the factors to determine whether an invention falls within the excluded items in Article 52(2) EPC. Thus, although the language “technical character” had vanished from the EPC, Germany’s technical character approach was revived in the rule and Guidelines for Examination for the EPC. Unfortunately, the term “technical character” has not been defined due to difficulties. Instead, defining the terms “technical” and “technical character” has been left to the EPO Boards of Appeal to construe and develop. Therefore, the scope of patentable subject matter has gone through a transition regarding the interpretation of the term “technical character.” The first opportunity to apply the new standard, “technical character” test, came in 1987 in T 208/84(In re Vicom).
2. Case Law in the EPO for Programs for Computers & Methods for Doing Business (Computer-Implemented Inventions)
The basis for EPO practice relating to patentable subject matter for computer-implemented inventions is the decision in T 208/84 (In re Vicom, 1987). The Vicom decision set forth the principle governing patentable subject matter for computer-implemented inventions. The claimed inventions in Vicom were directed to methods for digitally processing images characterized by a digital filter. By digital filtering, the number of calculations for data processing could be reduced, and then the processing speed substantially increased compared to the prior art.
The Board held that “even if the idea underlying an invention may be considered to reside in a mathematical method[,] a claim directed to a technical process in which the method is used does not seek protection for the mathematical method as such.” From the Board’s view, a technical process for a computer-implemented invention may mean a process which is “susceptible of industrial application” being “carried out under the control of program.” Interestingly, this holding is very similar to the holding in Diehr, in which the United States Supreme Court noted that Diehr sought to patent an industrial process for the molding of rubber, and had not attempted to patent an un-patentable law of nature or mathematical algorithm. Then the Board ruled that “a claim which can be considered as being directed to a computer set up to operate in accordance with a specified program … for controlling or carrying out a technical process cannot be regarded as relating to a computer program as such.”
Further, like Flook and Diehr, the Board emphasized that an invention should not be excluded from patent protection by the mere fact that a computer program is utilized for implementation, but should be considered as a whole.
Although the reasoning seems to partly mix the eligibility inquiry with that of patentability, after Vicom, a computer-implemented invention was patent eligible if its use was directed to some technical process, possessing industrial applicability that was implemented by computer programs. Thus, Vicom, as well as Diehr in the United States, opened the door to granting a computer-implemented invention a patent.
The next leading case in Europe is T 26/86 (Koch & Stezel, 1988), in which the Board examined whether an X-ray apparatus incorporating a data processing unit operating in accordance with a routine was patent eligible. The Board found that there was no constant technical interaction between the program and the X-ray apparatus, but there was a technical effect and the invention was therefore patent eligible. In reasoning, following Vicom, the Board stated that “an invention must be assessed as a whole.” Then the Board held that if an invention used both technical and non-technical means, “the use of non-technical means did not detract from the technical character of the overall teaching,” and the invention using technical means should be patent eligible. The Board therefore indicated that inventions consisting of a mix of technical and non-technical elements did not prohibit it from being patented, and that it is irrelevant “to weigh up the technical and non-technical features in a claim in order to decide whether it relates to a computer program as such.”
The Board supported Vicom (also Diehr) when it emphasized that an invention should be assessed as a whole, and clearly stated that considering the claimed non-technical and technical elements individually is irrelevant in a judgment of patent eligibility. In this sense, the holding seems to correspond with Diehr and Alappat, which virtually denied the Freeman-Walter-Abele test, even though the second prong in the Freeman-Walter-Abele test required the “as a whole” consideration, the first prong required to decide whether a clamed invention recites a mathematical algorithm.
In 1994, the Board further pushed forward to expand the scope of patent eligibility. In T769/92 (Sohei), the subject of the invention was a computer system and method for simplifying business management by allowing operators to input data relating to either inventory management or financial management into a single user interface (a “transfer slip”) and by processing that data and transferring it to the appropriate files. The claimed financial and inventory management would generally fall under “methods for doing business.” Since a system claim and method claim differ only by their categories, the Board adopted the same approach and considerations. Essentially, the invention was directed to a user interface comprising a single transfer slip displayed on a display screen of the computer system through which a debit item and credit item ran, and therefore it was considered to be a mix of technical elements (computer hardware) and non-technical elements (functional features implemented by software).
Relying on T 26/86 and T 38/86, the Board of Appeal noted that when a non-patent eligible (e.g. mathematical, mental or business) method is implemented by running a program (non-technical) on a general purpose computer, the fact alone that the computer consists of hardware (technical) could render the method patent-eligible under Articles 52(2), (3) EPC, if said software’s contribution to that computer (art) can be found either in a technical problem (to be) solved, or in a technical effect achieved by the solution (technical considerations). Also the Board found that such a mix may not be excluded from patentable subject matter, when technical considerations concern the particulars of its implementation, because technical consideration implies the occurrence of a technical problem to be solved (Rule 27 EPC ) and technical features (Rule 29 EPC ) solving that technical problem, may be regarded as resulting in a technical contribution to the art. Although each hardware unit recited in a claim was a part of a conventional general purpose computer, the Board found that the claimed five files “caused the processing unit to perform the … five functions would clearly require technical considerations.” Further the Board indicated that the claimed restriction introduced technical effects, and therefore the claimed system and method were inventions in the meaning of Article 52(1).
Since the technical contribution approach is used to judge an inventive step, the reasoning seems to confuse the assessment of patent eligibility with that of inventive steps. Importantly, however, the Board set forth the “technical consideration” approach to assess the patent eligibility of a mixed invention (technical and non-technical aspects), and confirmed that patent-eligibility is never reduced simply because the claim includes a feature which itself is excluded from patentable subject matter, as in the present case features referring to management systems and methods which may fall under “methods for doing business” under Article 52(2) EPC. Accordingly, the holding remarkably expanded the scope of an invention in the meaning of Article 52(1), since even if a claimed invention which can easily be classified as a method for doing business, like Sohei, has a great chance to be patented.
The trend narrowing the meaning of “as such” was not stopped, but rather continued. In 1999, two remarkable decisions were issued for the EPO Technical Board of Appeal, T1173/97 and T935/97 (IBM cases). These cases substantially correspond to In re Warmerdam, In re Lowry and In re Beauregard in the United States, and paved the way for patenting programs for computers. In these IBM cases, the Board confronted the issue whether “a computer program product” and “a computer readable medium having a program recorded thereon” were patentable subject matter. First, the Board construed the meaning of “as such” in Article 52(3) EPC by noting that the legislator did not want to exclude from patentability all programs for computers, but only programs for computers “as such.” Then the Board addressed whether requiring “technical character” for an invention is generally accepted in the context of the EPC in light of Rules 27(1) and 29(1) EPC, and whether “programs for computers as such … may be interpreted to mean that such programs are considered to be mere abstract creations, lacking in technical character.” Accordingly, the Board held that if programs for a computer have technical character, they must be regarded as within the scope of the patentable subject matter. Next, the Board was consulted to define the meaning of the “technical character” of programs for computers. Since the programs for computers inevitably result in physical modifications of the hardware, such as electric currents, the Board found that the further technical effects, which may be known in the art, but which go beyond the normal physical interactions between program and computer, are required to be the subject matter of a patent. Finally, regarding a computer program product, the Board applied “the further technical effect” to determine the patentable subject matter of claimed inventions, and then found that a computer program product possesses the required technical character because it has the potential to produce a further technical effect.
Before the IBM cases, a computer program was not regarded as patentable subject matter regardless of possessing a technical character. Thus, a computer program product was not excluded from patentable subject matter “if the program, when running on a computer or loaded into a computer, brought about, or was capable of bringing about, a technical effect which went beyond the normal physical interactions between the program (software) and the computer (hardware) on which it was run.” As a result, seen from the enforcement perspective, software business, to which the mainstream business of most IT companies began to shift, obtained a great benefit from the IBM decisions. On the other hand, as the appellant argued, the USPTO and the Japan Patent Office (JPO) already treated a computer program product as patentable subject matter at that time. Therefore, the decisions in the IBM cases took a valuable step toward harmonization in the examination of computer software inventions. Following the IBM decisions, it became accepted practice of the EPO that software itself could be claimed directly, and it continues to be the practice of the EPO.
In addition, the IBM cases influenced the practice in the EPO of the so-called “contribution approach,” which permitted some prior art to be taken into account when determining whether the subject matter was excluded under Article 52(2) or (3) EPC. In short, around the 1990’s, the practice sometimes mixed up the assessment of novelty and inventive step and that of patentable subject matter. The Board mentioned, however, that the practice was inappropriate for examining the subject matter of a patent as follows: “Determining the technical contribution an invention achieves with respect to the prior art is therefore more appropriate for the purpose of examining novelty and inventive step than for deciding on possible exclusion under Article 52(2) and (3) EPC.”
After the IBM cases, a more significant shift in approach was brought by the Pension Benefit decision (T 931/95) in 2000. In this case, the Board struggled with patent eligibility of inventions relating to methods for doing business. The subject matter was directed to a method and apparatus for controlling a pension benefit program by administrating at least one subscriber employer account. Regarding the method claim, the Board found that methods involving only economic concepts and practices of doing business were not an invention in the meaning of Article 52(1) EPC, because the claimed invention does not go beyond a method of doing business as such. The Board, however, did not take a clear position on “[a] feature of a method which concerns the use of technical means for a purely non-technical purpose and/or for processing purely non-technical information,” by saying that “such a feature does not necessarily confer technical character to [that method].” On the other hand, the Board noted that since having technical character was an implicit requirement of the EPC to be met by an invention in order to be an invention in the meaning of Article 52(1) EPC, if the method had a technical character, it still might be a method for doing business, not a method for doing business as such.
As for the apparatus claim, the Board distinguished it from a method claim relying on the fact that the language “apparatus” ‘in the sense of “physical entity” or “product”’ is not recited in Article 52(2)(c) EPC. Then the Board concluded that claiming such a physical entity or product in the form of an apparatus inherently possesses physical features which may be technical features. Eventually, the Board held that “an apparatus constituting a physical entity or concrete product suitable for performing or supporting an economic activity,” provides physical features (technical features) and thus is patent eligible under Article 52(1) EPC.
Applying these rules, the Board found that while a method claim was not an invention under the meaning of Article 52(1) EPC because it related to a method of doing business as such, an apparatus claim was considered as having technical features. The apparatus claim, however, failed to meet the inventive step requirement, since it was viewed as being different from the lay prior art in an economical field, that is a non-technical field, and hence there was no technical contribution provided by the distinguishing features of the invention. Thus, although both a method claim and an apparatus claim were examined equally in Sohei, in the present case, the Board treated them differently and explicitly made a favorable judgment for the apparatus claim. In other words, an apparatus for doing business is nearly automatically conferred patent eligibility. As a result, the Board decision would undermine the exclusion of business methods in the case of claiming an apparatus (physical entity or product).
Assuming the present case were analyzed on the ground of the machine-or-transformation test in the United States, it would be difficult for both a method claim and an apparatus claim to avoid rejection under 35 U.S.C. § 101. Although both claims include physical means, since these means are claimed abstractly, they are not deemed to impose meaningful limits on the claim scope. At first glance, the inventions in the present case seem to be an abstract idea itself. Moreover, an “article” under the transformation test is not found in the present case.
Although the decision in Pension Benefit has yet to fully endorse computer-implemented inventions, in Hitachi (T258/03, 2004), the Board expressed a favorable opinion for the patentability of computer-implemented inventions. In this case, the inventions at issue were a method, an apparatus and a computer program for automatic auctions. Regarding the apparatus claim, although the Board followed the established case law, in which a mix of technical and non-technical features may be an invention within the meaning of Article 52(1) EPC, the Board seems to abandon examining patent eligibility of such a mixed invention. Because it is difficult to distinguish technical from non-technical features in a claim, and “such technical features may be easier to identify within the framework of the [inventive step] examination,” the Board addressed that it is practical to regard the mixed features as patent eligible under Article 52(1) EPC. Then, without considering whether physical features, such as a server computer, client computers and a network, are “suitable for use in a particular field” as Pension Benefit did, the Board simply stated that the apparatus claim is patentable subject matter since it comprises physical features which are qualified as clearly technical features. Thus, in Pension Benefit, at least merely putting a general purpose computer in a claim would be regarded insufficient to render patent eligibility, but the present case suggests that including a physical feature in a claim would be enough so as to be an invention within the meaning of Article 52(1) EPC.
Regarding a method claim, the Board addressed “whether technical character is conferred to a method using technical means for a purely non-technical purpose,” and it also went beyond the Pension Benefit decision. The Board pointed to the Pension Benefit opinion as confusing the patent eligibility inquiry with novelty or inventive step inquiries, and held that technical character “may be conferred to a non-technical activity by the use of technical means,” as well as “implied by physical features of any entity or the nature of an activity.” Significantly, the Board admitted that the above holding is such a broad interpretation of the statute that conventional activities, “such as the act of writing using pen and paper,” could be qualified as technical character. Then the Board suggested that when computer-implemented inventions have mixed features, an appropriate inquiry is not the technical character (patent eligibility), but the technical consideration (inventive step).
Thus, the Pension Benefit and Hitachi cases limited the scope of exclusions under Article 52(2) EPC by shifting the assessment of technical character to the technical contribution approach under Articles 52(1) and 56 EPC. As a result, it substantially removed almost all subject matter relating to computer programs and methods for doing business from the exclusions. Even compared to the United States after Bilski, the Hitachi case showed a tolerant attitude toward computer-implemented inventions. Regarding apparatus claims, contrary to Hitachi, the machine-or-transformation test emphasizes the importance of the patent eligibility inquiry. In the present invention, since a usual Internet system is simply added to the claim and there is no transformation of an article into a different stage, the apparatus claim would be denied under § 101. As for the method claim, the present case is remarkably similar to the Cybersource case issued after Bilski, in which the district court found that “over the Internet” is a meaningless limitation and it rejected patent eligibility. Therefore, the present claim would suffer the same fate.
The Board’s recent holding in Microsoft (T 424/03 (2006)) further increased the tendency against requiring physicality in order to import patent eligibility to a claim. The subjects of the invention were a method and a computer program implemented in a computer system that has a clipboard for performing data transfer in a plurality of clipboard formats. In accordance with established case law (T 258/03), the Board found that a computer system including a memory (clipboard) was a technical means, and consequently that the claimed method has technical character. Furthermore, the Board stressed that “a method implemented in a computer system represents a sequence of steps actually performed achieving an effect, and not a sequence of computer-executable instructions (i.e. a computer program) which just have the potential of achieving such an effect when loaded into, and run on, a computer.” Thus the Board differentiated a computer-implemented method from a computer program based solely on a claim category, and concluded that such a computer-implemented method would never become a computer program as such. As for a computer readable medium claim, also citing T 258/03, the Board held that since the subject matter of the claim relates to a computer readable medium, which is a technical product, it possesses technical character.
The Board’s emphasis seems to provoke new controversy over computer-implemented inventions. This holding would suggest that only claiming a computer-implemented method, not a computer program, can avoid rejection under Article 52(2) and (3). However, since the nature of a program is a series of instructions for a computer and is supposed to be run on the computer, the degree of potentiality of achieving an effect would be as high as actual achievement. There are not and should not be any substantial differences between a computer-implemented method and a computer program at all. On the other hand, regarding a computer readable medium claim, the holding is substantially the same to stating that merely adding the term “a computer readable medium” to a claim would automatically render a patent eligibility without imposing meaningful limitations on the claim scope.
The holding obviously overemphasizes the claim form, category, and disparages the core of an invention. As a result, even meaningless claimed limits can import patent eligibility into the claimed invention. In this regard, the Board takes a completely opposite view of the machine-or-transformation test in the United States, which asks for imposing a substantial limit to the claim scope.
Compared to the United States approach, the machine-or-transformation test, the European approach has drastically narrowed the scope of exclusions under Article 52(2). Due mainly to the difficulty to separate technical features from non-technical features, Articles 52(2) and (3) gradually become a dead letter. Instead, the Board and examiners began to place more emphasis on Article 56 (inventive step). However, the established case law has required an invention not to be either abstract and/or non-technical so as to be granted a limited-term monopoly. Therefore, even if a mix (technical and non-technical features) computer-implemented invention would at least possess technical features, the inquiry whether the invention is so concrete that it is suitable for patent protection should be considered. Unlike the Pension Benefit, Hitachi and Microsoft cases, putting meaningless physical limits on a claim does not and should not be deemed to convert non-technical/abstract into technical feature/concrete.
In the meantime, on October 24, 2008, diverging decisions of the EPO's Boards of Appeal have created uncertainty, “EPO President Alison Brimelow has referred a number of questions to the Enlarged Board of Appeal of the EPO (EBoA) in relation to the patentability of programs for computers under the [EPC].” The case, G3/08 has pointed out the following four questions.
First, should whether or not the claimed computer program falls within the scope of the patentable subject matter be determined by whether it has a technical character (T 1173/97, IBM)? Or should it be determined by whether the invention is claimed in the form of a computer program or a method implemented in a computer system (T 424/03, Microsoft)?
Second, there are inconsistencies regarding the requirement for patent eligibility of computer programs. Although programs for computers have been considered to be a type of method claim, a method claimed in the form of a computer implemented method requires only technical means in order to be regarded as having technical character (T 258/03, Hitachi), while the same method claimed in a computer program requires a further technical effect, which must go beyond the normal technical effects (T 1173/97, IBM).
Third, must a claimed feature cause a technical effect on a physical entity in the real world in order to contribute to the technical character of the claim? If the answer is yes, is it sufficient that the physical entity be an unspecified (general purpose) computer (T 163/85, T 190/94)? On the other hand, if the answer is no, can features contribute to the technical character of the claim, when the only effects to which they contribute are independent of any particular hardware that may be used (T 125/01, T 424/03)?
Finally, does the activity of programming a computer necessarily involve technical considerations? If the answer yes, do all features resulting from programming thus contribute to the technical character of a claim (T 1177/97, T 172/03)? If the answer is no, can features resulting from programming contribute to the technical character, only when they contribute to a further technical effect when the program is executed (T 833/91, T204/93, T 769/92)?
The answers to these four questions have not yet been provided. Even whether the answers will be published is still unknown.
3. Recent Movements in Germany and the United Kingdom
As shown in section III. B. 1, the EPO practice, which basically requires an invention to possess technicality to be patent eligible, stems from the German approach. The German courts (the Federal Court of Justice (FCJ) and Federal Patent Court (FPC)) have improved their approaches under the German Patent Law 1981 § 1 in virtually the same manner as Article 52 EPC. The same core doctrine as the EPO was seen in the Logic Verification case (1999): the essence or core of patent eligibility is not a teaching of a systematic activity to use controllable natural forces anymore, rather to be “technical;” the technical character of the invention must be determined on the basis of the whole subject matter; and technical character can be present if technical considerations are necessary to solve the problem. After this case, Germany entered a pro-patent period.
During that period, a computer was treated as being generally technical, and providing further technical effect was considered irrelevant for assessing the technical character. However, the pro-patent era ended with the Search for Faulty Character Strings case announced in 2001, in which the prominent features that solve a concrete technical problem were required to be patent eligible. Thus, the German court eventually took the same position as the IBM cases from the EPO. The same the EPO, the German courts have struggled with the scope and interpretation of the term “technical,” but since the EPO approach was based on Germany’s, both approaches show a great similarity.
The U.K., meanwhile, had engaged in the “manner of new manufactures” approach until it was abandoned in 1977. Thereafter, the United Kingdom Intellectual Property Office (UKIPO) has struggled with following the new technicality approach, the EPO approach, to assess patent eligibility of computer-implemented inventions. However, the UKIPO is bound by the precedents set by the courts of the United Kingdom. This could often cause disparity between when the EPO and UKIPO will grant a patent.
Practice in the U.K. has relied upon the U.K. Court of Appeals judgment in Fujitsu (1997). In this case, the judge followed the previous U.K. cases ( Merrill Lynch and Gales ) and the EPO cases (Vicom) and applied a “technical contribution” approach to determine whether the invention falls within the nonstatutory subject matter: the claimed invention must provide technical contribution. However, the outcome seems to go beyond these precedents to some extent by stating that if an invention, which produces no more than a new computer program, would not produce technical contribution at all, and would be excluded from patentable subject matter. Thus, the UKIPO practice toward computer-implemented inventions was basically in line with that of the EPO at that time, although the inconsistency began to be seen. However, as the U.K. Court of Appeal noted later on, when the EPO shifted to its recent more restrictive stance to Article 52(2)(c) EPC exclusions in Pension Benefit, Hitachi and Microsoft, the UKIPO still continued to adopt the “technological contribution” approach. This was because UKIPO had to follow its precedent in Fujitsu. As a result, there was a clear inconsistency in the way in which the UKIPO and the EPO interpreted the language of the EPC.
In the meantime, in 2006, two cases regarding disputes of exclusions of statutory subject matter reached the U.K. Court of Appeals and were considered in a single decision, Aerotel/Macrossan. In Aerotel, the court reaffirmed its support of the traditional U.K. approach (technical contribution approach) relied on Fujitsu to exclude subject matter. However, following two judgments in the Patents Court (the lower court), the UKIPO published a practice notice announcing that it would not reject claims directed to a computer program or a program on a medium merely because of their claim form. One of these Patents Court’s judgments is in Symbian Limited v Comptroller General of Patents. This case was an appeal by the UKIPO against a decision by the Patents Court to reverse the UKIPO’s refusal of Symbian’s computer-implemented application relating to accessing data in a dynamic link library in a computing device.
On October 8, 2008, the Court of Appeals issued its decision. In this case, the court emphasized the importance to diminish the inconsistency between the U.K. practice and the EPO practice. After scrutinizing U.K. and EPO case law, the court decided to follow Merrill Lynch, Gales (the U.K. cases), Vicom and IBM (the EPO cases), but not Fujitsu and Aerotel. Further, although the court admitted the impreciseness of the boundary line between what is and what is not technical contribution, it found that these four cases (Merrill Lynch, Gales, Vicom and IBM) provide the most reliable guidance. The court warned, however, that because of “the multifarious features of computer programs and the unpredictable developments which will no doubt occur in the IT field, we believe that it would also be dangerous to suggest that there is a clear rule available to determine whether or not a program is excluded by art 52(2)(c),” and appealed for the importance of case by case consideration (both facts and features). Applying the technical contribution approach, the court held that the present invention was not excluded from patentable subject matter.
Although the Court of Appeal’s judgment did not go as far as adopting the recent EPO approach to computer-implemented inventions, importantly it noted that the U.K. and EPO approaches should give the same result. This judgment suggests that the U.K. stance on patentable subject matter for computer-implemented inventions, even if it does not necessarily follow the same reasoning, it should be more in line with that of the EPO. Regarding the recent EPO approach seen in Pension Benefit, Hitachi and Microsoft, the EPO Board is now confronting the divergence of its decisions in G 03/8. Therefore, the outcome of G 03/8 will be very influential in not only the EPO, but also the in U.K. The impact of the present decision and G 03/8, if issued, on the UKIPO and the U.K. courts remains to be seen.
C. Patent Eligibility in Japan
1. The Japan Patent Act and Examination Guidelines
Japan is a civil law country, and hence has fewer court decisions to provide guidance as common law countries do. Therefore, relevant patent law is codified in the form of statutes and administrative guidelines. Japanese patent law, which is generally based on German patent law, is quite similar to European countries’ patent laws. The improvement in Japanese patent law, however, has been strongly affected by the United States patent experience. As a result, the Japanese practice for computer software and business method inventions seems to be somewhere between these two other major patent law regimes.
Characteristically, unlike most other jurisdictions including the United States and European countries, the Japan Patent Act (JPA) defines “invention.” Article 2(1) states that an “invention” means the highly advanced creation of technical ideas utilizing the laws of nature. This definition is synonymous with the definition introduced in Kohler in 1908. Article 2(1), however, did not exist in the patent act until 1959. During the revision of the Japan Patent Act in 1959, the provision was introduced because of the policy consideration to show a simple and clear definition to the public and to reduce conflict.
The language “utilizing the laws of nature” in Article 2(1) was referenced partly from two cases decided by the Tokyo High Court in 1950 and 1953. In both cases, the subject matter was directed to the method of creating a telegram code by combining characters, numbers and marks, and the court indicated that even if the invention might contribute to business and might be scientifically excellent, it neither utilized a machine nor adopted methods utilizing natural forces. ,  The court held that such a claimed method was not an invention under the meaning of the Japan Patent Act 1921. Seemingly, the holding could be interpreted that including a machine in a claim would be enough to be regarded as utilizing natural forces and then it would be patent eligible.
In 1956, however, the Tokyo High Court explicitly denied the patent eligibility of an apparatus for posting advertisements on telephone poles because the invention did not utilize natural forces. Therefore, by introducing the term, “utilizing the laws of nature,” the Patent Act of1959 announced its intention that the requirement be adopted regardless of claim categories. Furthermore, requiring utilization of laws of nature, the language appeared to exclude a mental process and a law of nature itself from statutory subject matter. Moreover, regarding the term a “creation of technical ideas” in Article 2(1), the courts have interpreted it to be constituted of three factors: repetitive practicability, concreteness and objectiveness. These factors would have an important implication for assessing the patent eligibility of computer software-related inventions later.
Since Article 2(1) does not define the invention that should be protected by patent law other than a general definition, the introduction of Article 29 is a substantial provision for patenting requirements. It requires being at least an “invention” which is industrially applicable in order to obtain patent protection, following Article 1 of German Patent Act, “die eine gewerbliche Verwertung gestatten.” As a result, to be granted a patent, a creation of technical ideas, utilizing a law of nature and industrial applicability are required. Usually, however, computer software inventions satisfy industrial applicability. Consequently, utilizing a law of nature and being a creation of technical ideas might be more important requirements for patent eligibility. In short, Article 2, rather than Article 29, plays an important role for assessing patent eligibility of computer software-related inventions in Japan.
Providing additional guidance, the Japan Patent Office (JPO) has issued and revised the Examination Guidelines for Patent and Utility Models in Japan since 2000. Before 2000, the guidelines focused on each technical field. As far as the computer software-related inventions are concerned, the examination guideline for computer programs was issued on December 26, 1975 (the 1975 guideline). The 1975 guideline applied to computer program-related inventions, which claimed a procedure corresponding to the substance of computer programs in the form of at least one claim element. According to this guideline, if a formula used for producing an intended result did not utilize a law of nature, such as laws of nature themselves or arbitrary arrangements, a technical idea based on the formula could not be statutory subject matte. On the other hand, if a formula utilized a law of nature, a technical idea based on that formula could be. In other words, if a cause-and-effect relationship achieved by an invention utilizes a law of nature, the invention would fall within the scope of patentable subject matter. Relying on this logic, the inventions not only in Diehr but in Flook might be regarded as patent eligible because each cause-and-effect relationship relied on by each mathematical algorithm seemed to contain a law of nature. Notably, however, the 1975 guideline clarified that a computer program-related invention was only patent eligible as a “process” claim. Therefore, a product claim of a computer program-related invention was not allowed, and a computer program as such was denied patent eligibility due to its significant abstractness.
Since the 1975 guideline limited the scope of patentable subject matter to only a process claim, the limitation became unreasonable as microcomputer-applied technology exploded. To respond to the changing social environment, in 1982 the JPO published the Examination Guidelines for Inventions Relating to Microcomputer-Applied Technology (the 1982 guideline), which declared that not only a process, but a product claim relating to microcomputer-applied technology could be patentable. Under the 1982 guideline, if the means for practicing a function corresponding to computer programs was claimed in the form of a component of an apparatus, the apparatus could be statutory subject matter.
In reality, however, some inventions which might go beyond the 1975 and 1982 guidelines were granted patents. For instance, the invention in Benson was filed in Japan and then patented. Although the invention did not claim a form of a process, but rather an apparatus, the substance of these inventions were almost identical. Considering the implication in Benson that the U.S. Supreme Court would reach the same conclusion regardless of a process claim or a product claim, the invention filed in Japan would have been rejected patent eligibility in the United States. Also, even under the 1982 guidelines in Japan, because the claimed apparatus did not claim the means for practicing the conversion from one binary form to another binary form, rather it simply claimed the existence of shift register, it might be deemed nonpatentable subject matter. Thus, although variability was observed to some extent in the application of the guideline, Japan has struggled with the abstractness of computer software-related inventions and was looking for clues in physical entities and technicality just like the United States and Europe.
In 1988, the proposed guideline, Examination Method of Computer Software Related Invention, was issued, and it was released in 1993 as a part of the new Examination Guidelines for Specific Technological Fields (the 1993 guideline). The 1993 guideline further enlarged the scope of patentable subject matter in order to respond to explosion of computer software-related technology. The 1993 guideline, like the United States and European practices, declared that a claim should be considered as a whole in the judgment of statutory subject matter. Then it instructed that either utilizing a law of nature in information processing by software or utilizing hardware resources was required for the computer software-related inventions to be considered patentable subject matter. A computer program as such was still outside the patent protection since it was not a “technical idea.” Regarding claim category, the 1993 guideline suggested that describing a claim in a form of product or process did not matter in determining the utilization of a law of nature. However, at that point, computer programs stored on a tangible medium were not accepted as statutory inventions in Japan as they were in the United States and Europe.
After the issuance of the 1993 guideline, however, especially in the United States, the necessity of patent protection of computer programs stored on a tangible medium had emerged in the world, such as in In re Lowry and In re Beauregard. Considering international harmonization and the desire from the emerging hardware and software industries seeking patent protection for computer programs, in 1997, prior to Europe, the JPO promulgated that computer-readable medium with computer programs could be protected under the 1997 Examination Guidelines for the Computer Software-Related Inventions. Further, due to rapid development of Internet technology, the business models of most hardware and software companies shifted from hardware to software, a computer program itself. At the end of the twentieth century, the boom in IT business further pushed the patent protection of computer software inventions. Claiming a computer program itself was allowed in the United States. when it is drafted in the form of a means-plus-function claim, and it have been also accepted by the EPO as one of the types of computer-implemented inventions. Especially in the EPO, since a computer program is supposed to be used on a computer, claiming hardware resources is not strictly required unless a skilled person cannot recognized the existence of them. In the meantime, in 2000, the JPO finally announced that computer programs not embodied in a tangible medium can also be patentable subject matter as a “product” invention in the Examination Guidelines for Computer Software-Related Inventions (the CS guideline) which is one of the parts of the latest Examination Guidelines of 2000 (the 2000 guideline). Furthermore, the CS guideline refined the test to determine the patent eligibility: where information processing by software is concretely realized by using hardware resources, the software is deemed to be “a creation of technical ideas utilizing a law of nature,” then it falls within the patent subject matter. The test, like in the United States and Europe, requires “as a whole” consideration.
While the 2000 guideline began to treat a program claim as a product claim, it was still unclear to what extent the Japan Patent Act covered intangible entities because the Japan Patent Act was legislated in 1959 based only on tangible entities. Unlike the United States and the Europe, since the Japan Patent Act is a special law of the Japanese Civil Law, in which a product has been defined as a tangible entity, it had been desired to specially define the meaning of a “product” in the Patent Act. Then, in 2002, the Patent Act Article 2(3) was revised so as to stipulate that a product includes a program, an intangible entity. At the same time, there was another significant change in Article 2(3). From the aspect of enforcement, providing a program via the Internet was defined as one of the forms of using the claimed invention, which is one of the actions that constitutes patent infringement. Thus, in 2002, considering the change in business models and international harmonization, the Japanese government took a significant step toward protecting computer software.
Under the CS guideline, a computer software-related invention means an invention that needs software for practicing the invention. If a computer software-related invention requires special judgment or treatment in examining, the CS guideline is applied. Otherwise, the general part in the 2000 guideline is applied. As examples of computer software-related inventions that do not require special judgment or treatment, the CS guideline includes claimed inventions directing to mathematical methods, mental activity or arbitrary arrangements. Even though a business method invention includes computer software in a claim, it would be examined under the 2000 guideline when it is deemed to be mainly directed to an arbitrary arrangement itself. In sum, a computer software-related invention is not always analyzed under the CS guideline merely because of using software, and as a result it is not always required that there be cooperative working between software and hardware resources so as to realize the invention. In this sense, not only the CS guideline, but also the general and principle part in the 2000 guideline are closely relevant to determining the scope of patentable subject matter.
Where the CS guideline is applied to a computer software-related invention, whether information processing by software is concretely realized by using hardware resources is examined. Article 2(1) requires an “invention” to utilize a law of nature and be a creation of technical ideas. According to the drafters of the CS guideline, however, in case of computer software-related inventions, there is no need to decide whether such inventions utilize laws of nature in detail. If computer software itself is regarded as a creation of technical ideas, it is not the same as mathematical methods, mental activity or arbitrary arrangements any more. Rather it becomes a specialized algorithm for a computer that enables the computer to process information by operating hardware resources in it, utilizing a law of nature.
Therefore, the heart of patent eligibility for a computer software-related invention, which is applied to the CS guideline, would be whether an invention is a creation of technical ideas. In this regard, emphasizing the technical idea of software may be reminiscent of Alappat, in which the Federal Circuit implied that certain software could transform a general purpose computer into a special purpose computer. As discussed above, precedent in Japan has suggested that “creation of technical ideas” is satisfied when practicability, concreteness and objectiveness are satisfied. In case of a computer software-related invention, however, there is little doubt that a computer software-related invention meets practicability and objectiveness if it is concrete so as to be recognized that software is realized on a computer. Consequently, “concreteness” is the only substantial requirement for patent eligibility of a computer software-related invention, and the test whether information processing by software is concretely realized by using hardware resources is important.
The language “using hardware resources” sometimes might receive undue focused on, and causes misunderstanding that claiming in detail some hardware components, such as CPU, memory and display, is a requirement. However, rather concretely claiming processing information is a substantial requirement, and by doing so, it might be inevitable to identify the interrelationship between software and hardware resources to some extent. If identifying the interrelationship so as to obviously understand that the processing information premises on using hardware resources, then information processing by software is considered as being concretely realized by using hardware resources.
Regarding business method inventions, the number of applications relating to business methods sharply rose from around 2,000 in 1999 to more than 12,000 in 2000, due mainly to a trend of IT business and the State Street Bank decision in the United States in 1998. In the meantime, since almost all inventions characterized by business methods could be considered as computer software-related inventions, the JPO announced that business methods inventions should be examined as other computer software-related inventions and not be treated specially. Even before State Street Bank, however, business method-related inventions were examined based on the Examination Guideline and some of them were granted patents. Furthermore, the judgment by the Tokyo High Court in 1950 implied that at least there was no so-called business method exception, and a prior Tokyo High Court judge suggested in his book that even though software for business methods was not patent eligible, it could be when it was connected to hardware resources. Thus, business methods have never been treated special, but have been examined equally to other inventions. Therefore, certain business method inventions have been regarded as being patent eligible and continue to be, as far as they meet the criterion set forth by the Examination Guideline.
Thus, the scope of the statutory subject matter in Japan has been significantly enlarged mainly by extending the allowable claim category. Unlike the United States and Europe, the criterion to determine patent eligibility for both computer software-related inventions and business methods inventions has not drastically changed. Rather, Japan has constantly emphasized utilizing laws of nature and creation of technical ideas, practicability, concreteness and objectiveness. Consequently, cooperative working between software and hardware resources has been required to be regarded as statutory subject matter in Japan.
2. Case law in Japan (Computer Software)
In 2006, the Intellectual Property High court (the IP high court) published its decision relating to patent eligibility of computer software-related inventions (Heisei 17 gyo ke 10698). The invention at issue was a method for managing users’ points, such as memorizing the amount of points and calculating points. Contrary to the applicant’s argument, the court held that although some hardware resources, such as a “network” and “point account database,” are included in a claim, claimed steps are able to be implemented by either a person or a computer since the subject of each claimed step is not specified. Then the court applied the 2000 guideline to the invention conducted by a person, and the CS guideline to the invention implemented by a computer.
As for managing points by a person, the court reasoned that “memorizing the accumulated points,” “receiving point” and “adding points” can be done by person, and noted that the present invention uses means, “network” and “database,” but considered as a whole, the invention was an arbitrary arrangement per se using those means as tools. Therefore, the court held that it does not constitute “a creation of technical ideas utilizing a law of nature.” On the other hand, in the case when a computer implements the invention, the court found that there is no direct indication of hardware resources in the claim. Although the terms “network” and “database” are recited, the court stated that these terms are not necessarily limited to utilizing a computer. As a result, the court concluded that using a computer may be assumed, but cooperative working between software and hardware so as to manage points is not observed at all.
The present case might be rejected under the machine-or-transformation test too, because merely putting the words “network” and “database” would not be regarded as imposing meaningful limits on the claim scope, and point data are also not considered as “articles” in the meaning of the transformation test.
Interestingly, in 2008, the IP high court confronted a patent eligibility issue for a Benson-like invention. In Heisei 19 gyo ke 10239, the invention was an apparatus for producing a shortened representation of a collection of bits by using the mathematical algorithm (a well-known hash function). The invention claimed some steps for calculating and recited one hardware resource, which was no more than an “apparatus.” The only issue was whether the claimed apparatus was a creation of technical ideas utilizing a law of nature. At first, the court confirmed the bedrock principle that mathematical algorithms as such cannot be patentable subject matter because they do not utilize laws of nature at all. Further the court found that using an existing computer for implementing the algorithm is no more than realizing the algorithm, and therefore the result would not be changed unless the creation of a technical idea is found out in the computer. To support its reasoning, the court expressed the fear, which may be reminiscent of Benson, that if the present invention is an invention in the meaning of Article 2(1) JPA, all mathematic algorithms would be the inventions. Then, referring to the CS guideline, the court held that since the invention does not claim any concrete circuit constructions to implement the algorithm on the computer, it does not add new creations to the existing computer at all: the essence of the invention is the mathematical algorithm as such, which cannot be patent eligible.
In contrast, the present invention was filed in the USPTO as an “effective universal hashing method,” which is almost identical to the Japanese claims except for the claim category, and it eventually issued on February 6, 2007. This result may suggest a divergence to some extent between the United States and Japan before Bislki.
In these two cases, citing the Examination Guideline, the IP high court affirmed the Board’s decisions that upheld the examiners’ rejection due to failing to meet the patent eligibility requirement. Within a year from October 2007, however, the court reversed the following three cases on the ground of the “as a whole” consideration (Heisei 19 gyo ke 10056 (Medical Bags), Heisei 19 gyo ke 10369 (Dental Therapy) and Heisei 20 gyo ke 10001 (Bilingual Dictionaries)) .
As for Medical Bags, the subject matter was a method for using a bag of medicine with cutoff line. The claim included four processes; first, printing information (such as a patient name, a medicine name) on a bag; second, putting the medicine into the bag; third, giving the bag containing the medicine to the patient; and last, cutting off the bag by the patient. Cutting off the upper part of the bag, in which personal information was printed, along with the cutoff line, the invention enables patients to protect their personal information. Like the JPO Board, the court found that who cuts off the upper part of the bag is an arbitrary arrangement. The court, however, noted that not the last step, but the first step, which utilized a law of nature is the essence of the invention so as to achieve the effects. Therefore, even though the invention includes arbitrary arrangements, considered as a whole it was deemed to utilize a law of nature, and thus is an invention in the meaning of the Article 2(1) JPA.
Also, the IP high court in Heisei 19 gyo ke 10056 expressed the same reasoning as the Medical Bags case. In this case, a claimed dental therapy system recited the means for deciding what dental therapy is necessary, and the means for making a treatment plan. However, after stating that these means include mental processes, the court found that the invention is not directed mainly to these mental processes, rather considered as a whole, it provides the technical means for aiding dental therapy, which contains a network server, communication network and a computer with a display placed in a treatment room. Then, the court concluded that the present invention is patent eligible.
In the last case, the subject matter was a method for referencing a bilingual (English and non-English) dictionary. In this case, since the method did not claim and does not necessarily require a computer, the 2000 guideline, not the CS guideline, was consulted. The method enables users to find the right English word without knowing a correct spelling by utilizing the nature of a person’s higher ability to discriminate a consonant than that of vowel sound. Therefore, the invention heavily relied on the inherent ability of a person. The court reasoned that whether a claimed invention falls within the scope of the patentable subject matter, it should be decided whether utilization of laws of nature is a central means of the creation of a technical idea aimed to resolve a problem. Patent eligibility should not be denied only because the invention contains a mental process or elements relevant to it. Applying the ruling, the court remarkably stated that since an inherent high ability to discriminate consonants provides a certain iteration effect and creation of a technical idea utilizing laws of nature is described as a central means of the invention, it is an invention under the meaning of Article 2(1) JPA.
If the present case were brought to a U.S. court or the EPO Board, the invention would be found difficult to be statutory subject matter in the United States, alternatively it would be statutory in Europe. In the United States, there is no doubt that it does not meet the machine test. As an article in the transformation test, consonant, vowel, accent marks, spellings and translation would be possible. Following the Bilski instructions, however, consonant, vowel, accent marks, spellings and translation are not deemed to be physical objects or substances, or an electronic signal representative of any physical object or substance. Rather they are closer to a data gathering step, which is classified as a non-article. Accordingly, the transformation test would not be satisfied. Thus, since “potential” computer-software inventions that are applied to the 2000 guideline do not necessarily claim hardware resources, they would be substantially examined under the transformation test in the United States. Because whether the invention utilizes a law of nature is a central question under the 2000 guideline, the scope of patentable subject matter would likely be broader than that of the transformation test.
On the other hand, under the EPO approach, whether the present invention possesses “technical” aspects seems be a tough question, because the invention does not explicitly claim a computer, it just states “suitable for computerized processing.” According to the EPO guideline, “the arrangement or manner of representation, as distinct from the information content, may well constitute a patentable technical feature.” In this case, therefore, the invention would be considered as such arrangements, and thus patent eligible.
As seen above, the court seems to suggest that the practice of “as a whole” consideration in the JPO is restricted to some extent. In the first two cases, there would be little doubt that the claimed arbitrary arrangements and mental processes are not the heart of each invention. In the last case, however, can utilizing an inherited ability be regarded as utilizing a law of nature as far as it produces a certain result again and again? If so, to what extent is the iterative nature required? Whether utilizing the inherited ability, which repeatedly provides a certain result is deemed to utilize a law of nature seems to be debatable. However, this case might trigger the emergence of the invention that utilizes an inherent nature of a human.
3. Case law in Japan (Business Methods)
The IP high court has not shown a clear stance toward the patent eligibility of business method inventions so far, but like the court’s previous dicta, the court made a short statement on August 28, 2008 against the JPO Board’s decision (Heisei 19 gyo ke 10327). In that case, the invention was directed to a method for e-commerce by selecting items in a computer display. The Board of the JPO found that although the invention uses a computer, it uses the computer as a mere tool for the claimed arbitrary arrangements and mental processes, and it as a whole is regarded as per se arbitrary arrangements and mental processes. After the Board’s rejection, the applicant appealed to the IP high court on the grounds of patent eligibility and inventive step. As for patent eligibility, the court stated no more than that it does not consider the Board’s decision legitimate. Then the court shifted to the inventive step inquiry.
Thus, although the reasoning of the IP high court remains unclear, the court seems to emphasize the “as a whole” consideration by taking into account the characteristics of the technology when the invention partly includes arbitrary arrangements and/or mental processes, as well as in the computer software cases mentioned above. In this sense, from the IP high court view, although the JPO Board applies “as a whole” analysis to an invention, it may still persist in the claimed arbitrary arrangements and/or mental processes. The impact of these decisions on the JPO remains to be seen.
- IV. COMPARISON AND ANALYSIS
Next, this paper will discuss three questions currently pending referral from the EPO, and then analyze the appropriate test to determine whether a computer software invention falls within the scope of the patentable subject matter.
A. Pending referral to the Enlarged Board of Appeal of the EPO (G 3/08)
First, should whether the claimed computer program falls within the scope of patentable subject matter be determined by whether it has a technical character (T 1173/97, IBM)? Or should it be determined by whether the invention is claimed in the form of a computer program or a method implemented in a computer system (T 424/03, Microsoft)?
The latter test could not withstand the criticism that it would be ridiculous for patentable subject matter to depend on the form of a claim, especially when there is no substantial differences between claimed inventions. Also, this test would conflict with the EPO guideline, which states that “the examiner should disregard the form or kind of claim and concentrate on its content in order to identify whether the claimed subject-matter, considered as a whole, has a technical character.” In the United States, denying patent eligibility merely because of a category or the form of claiming has been abandoned since State Street Bank denied the so-called “mathematical algorithm” exception and the business method exception. Even in Japan, there is no legal basis to deal with these claims differently after a program claim was allowed in 2000.
In T 424/03, the Board emphasized that “a method implemented in a computer system represents a sequence of steps actually performed and achieving an effect, and not a sequence of computer-executable instructions (i.e. a computer program) which just have the potential of achieving such an effect when loaded into, and run on, a computer.” In an ordinary way, however, a computer program is premised on being loaded into and run on a computer, and the phrase “a method implemented in a computer system” does not virtually add anything more than claiming “a computer program.” Needless to say, it is possible to claim “a method implemented in a computer system” without specifying technical means, such as a memory and a display, and also possible to claim “a computer program” with identifying specific technical means. Therefore, distinguishing the claim category of a computer-implemented method from that of a computer program just because of claiming “a method implemented in a computer system,” may misjudge the nature of computer-implemented inventions including computer programs. The statutory subject matter should be determined not by the claim form, but by substantial aspects, such as technical character, based on the established case law. Therefore, from the viewpoint of not only international harmonization, but also legality, whether the claimed computer program falls within the scope of the patentable subject matter should be determined by whether it has a technical character.
Second question is about inconsistencies regarding the requirement for patent eligibility of computer programs. Although programs for computers have been considered to be a type of method claim, a method claimed in the form of a computer implemented method requires only technical means in order to be regarded as having technical character (T 258/03, Hitachi), while the same method claimed in a computer program requires a further technical effect, which must go beyond the normal technical effects (T 1173/97, IBM).
Even though the technical character approach is an approach unique to Europe, both the United States and Japan basically have treated method claims and computer program claims equally. In the United States, the machine-or-transformation test is not only applied to method claims but also apparatus claims. Importantly, this test requires imposing meaningful limits on the claim’s scope and/or significant post-solution activity to import patent eligibility. Similarly, in Japan, the CS guideline announces that in case the information processing by software is not concretely realized by utilizing hardware recourses even if hardware resources are claimed, the claimed invention is not patent eligible. Thus, in these jurisdictions, computer software inventions cannot be patent eligible simply by citing technical means, such as a memory, regardless of either a computer implemented method or a computer program.
As the Board in the EPO indicated in T 1173/97, the substance of a computer program claim lies in the method that it is intended to carry out when being loaded into and run on a computer. It seems, therefore, illogical to distinguish between computer-implemented methods and computer programs, both of which will cause a method to be implemented. In addition, since these inventions are very closely linked to hardware resources (technical means), requiring no “further technical effects” could be nearly equal to conferring patent eligibility on computer program as such. Therefore, regardless of the category, a computer implemented method and a computer program should follow the case law established in T 1173/97 (IBM).
Third, must a claimed feature cause a technical effect on a physical entity in the real world in order to contribute to the technical character of the claim? If the answer yes, is it sufficient that the physical entity be an unspecified (general purpose) computer (T 163/85, T 190/94)? If the answer is no, can features contribute to the technical character of the claim, when the only effects to which they contribute are independent of any particular hardware that may be used (T 125/01, T 424/03)?
Although a technical effect is also a characteristic factor of Europe, “a technical effect on a physical entity in the real world” seems to be similar to the transformation prong in Bilski (must transform an article into a different state or thing). In Bilski, the Federal Circuit provided some examples of an “article”: any physical object or substance, or an electronic signal representative of any physical object or substance. On the other hand, the court stated that process claims of graphically displaying variances of data from average values, not specifying any particular type or nature of data, not specifying how or from where data was obtained or what the data represented and just gathering data are not an “article.” Thus, in the United States, the technical effect contributing to patent eligibility may not be limited to a physical entity in the real world and any particular hardware. In Japan, the CS guideline requires cooperative working between software and hardware resources so as to realize the invention, but it focuses less on the “effects” on the claimed physical entity. Instead, “effects” are considered at the stage of the inventive step inquiry, and the CS guideline addresses that effects such as quickness, processing a large amount of data, and obtaining uniform results cannot usually be considered an inventive step.
Considering this information society, requiring a technical effect on a physical entity in the real world might become an outdated idea. Requiring effects on any particular hardware, therefore, may be reasonable to exclude a computer program as such from statutory subject matter. However, to what extent technical effects can also be technical character should be determined by considering the purpose of the requirement and international harmonization.
B. Analysis and Suggestion of the Appropriate Test
Thus far, this paper has presented an overview of the tests for assessing patent eligibility created in each jurisdiction. Next, it will discuss the machine-or-transformation test in the United States as compared to the European and Japanese approaches.
The scope of patentable subject matter in the United States seems to have been relatively changeable compared to the other two regimes. One of the reasons for this may be the caused by lack of an explicit definition or bedrock concept regarding what an invention is in the meaning of United States patent law. While the EPO has persisted in “technicality” and the JPO has pursued “utilizing laws of nature” and “a creation of technical ideas” to draw the line between what is and is not an invention under each patent law, in the United States, there is little guidance to define and limit the scope of patentable subject matter, except judicially created exceptions. Rather, relying on Congressional intent, U.S. courts have enlarged that scope as if almost any subject would be patent eligible as far as it is “useful,” so as to catch up with the advent of useful new technology. Under the encouragement of this trend, “the USPTO has issued patents involving inventions from a broad range of disciplines, including a golf putt, teaching methods, and techniques of psychological analysis.” In the meantime, to response mainly to the Supreme Court’s opinions, the CAFC set forth the machine-or-transformation test.
On what does the Bilski court rely to limit the scope of patentable subject matter to the machine-or-transformation test? Are there any necessities to restrict the ambit of statutory subject matter in this way? Necessity may be caused by the nature of computer software, that is “abstractness.” The invention is inherently abstract compared to tangible property, but software inventions would be much more. That is, software itself does not necessarily require any connection with physical entities to be completed, except for running on a computer. Since abstract ideas have been eliminated from patentable subject matter in the United States, computer software inventions that are deemed to be abstract ideas should be excluded as well as other field inventions. As Judge Rader stated in his dissenting opinion in Bilski, being abstract ideas may be a sufficient reason for § 101 rejection in that particular case. However, under the useful, concrete and tangible result test, the abstract idea exception seems to be nearly slighted. Computer software inventions tend to be located near abstract ideas as such, so some sort of test to draw the line between what is and is not an abstract idea would be useful.
The Bilski test, however, is not new and “innovative.” In fact, this test showed up in various forms prior to State Street Bank, on which the Bilski test relied. In State Street Bank, however, Judge Rich marked the end to judicial efforts to judge “how much transformation is transformative” or “how much linkage and to what kind of machine serves the transcend abstractness.” After struggling with patent eligibility cases through the 1980s and early 1990s, the CAFC finally discarded the test which required defining degrees of transformation and machine connection. In this regard, Bilski revived such a troublesome task, and already it might cause instability and unpredictability in the law because of the difficulty to apply the test. For example, the ambiguity of the terms “transformation,” “meaningful,” “insignificant,” and “specific” was pointed out by Judge Newman and Judge Rader. Moreover, Judge Newman strongly dissents from the majority opinion, arguing that statutory history, Supreme Court and CAFC precedents do not support the majority ruling which “imposes a new and far-reaching restriction” on the invention suitable for patent protection. Although she scrutinizes the English Statute which was the origin of the United States Patent Act of 1970, the concurring opinion of Judge Dyk suggested that the statutory history may not be so decisive. Rather, as Judge Newman indicates, because the Bilski test relies only on the Supreme Court decisions, the fact that the machine-or-transformation test in Benson and Diehr was never treated as an exclusive test to assess patent eligibility seems to be lethal. Although the Bilski test might be a significant outcome toward setting a line between what is and is not an abstract idea, the legal basis for the line may be somewhat unclear.
As shown in the increasing number of § 101 rejections issued by the Board after Bilski, the machine-or-transformation test has resulted in a huge negative influence on computer software inventions so far, especially computer-readable medium and business method inventions. Are the Bilski test and thus trend reasonable compared to the approaches in Europe and Japan?
Liberally interpreted, the machine-or-transformation test comes closer to the tests adopted in the EPO and JPO. As for the machine prong, it corresponds to the technical means requirement in Europe. In Hitachi, even reciting technical means for a purely non-technical purpose would be statutory subject matter, regardless of the claim category. Further a general purpose computer could be such technical means. That is, imposing meaningful limits on the claim scope is not necessarily required to be statutory in Europe. In this sense, the European approach is fairly broader than that of the United States. The same would be true of a computer program and computer-readable medium claim. Following IBM, a computer program and medium claim would require further technical effects to be statutory. Seemingly this is nearly the same as the meaningful limits and/or significant post-solution activity in the Bilski test.
However, according to Microsoft, only claiming a computer-implemented method could be enough to be patentable subject matter. Thus, by changing the claim category, computer programs can easily pass the eligibility inquiry in the EPO. Moreover, there is no case law, statutes or rules such as Nuijten in Europe. The EPO guidelines only note that if the presentation of information has new technical features, there could be patentable subject-matter in the information carrier.” While the computer software embedded on a carrier or signal could be statutory in Europe as far as possessing new technical features, it would no longer be statutory in the United States. Consequently, the scope of patent protection against computer programs in the United States is now much more restrictive than that of Europe.
The machine test is also analogous to the requirement in Japan: cooperative working between software and hardware resources (information processing by software is concretely realized by using hardware resources). As seen in the term “cooperative” or “concretely realized,” imposing meaningful limits on the claim scope is also required in Japan. This requirement is applied regardless of the claim category. Accordingly, the United States machine test seems to be much closer to the Japanese test than that of Europe. In other words, the United States and Japanese approaches would be more restrictive than that of Europe.
What about a computer-readable media claim in Japan? According to the CS guideline, a patent seeking for “program signal(s),” “data signal(s)” and “an information transmission medium (e.g. a communication network)” cannot be rejected under the provision analogous to the 35 U.S.C § 112 (2), since they cannot be classified into a statutory category (a process or a product). The CS guideline, however, does not state that they are non-statutory subject matter. Thus, the Japanese approach is also quite different from the recent United States practice, which seems to treat such inventions as non-statutory subject matter. Significantly, since the U.S. practice attempts to eliminate a certain technical field (signals) from patentable subject matter, this may violate TRIPs Article 27, which states that patents “shall be available … in all fields of technology.” Now it is at least clear that the United States is one of the most difficult jurisdictions for a computer-readable medium claim to pass the patent eligibility examination.
As for the transformation inquiry, it seems to emphasize the results and/or effects on an article in a claim, rather than the claimed article itself. In this sense, this test would be similar to the “further technical effects” test in Europe. However, the further technical effects test is not applied to all field technology and the method category, but only to computer programs. Considering the EPO practice, the results and effects are more often assessed as a technical contribution in the inventive step (non-obviousness) inquiry (as shown in Part V). Therefore, although the interpretation of the term “further” would be instructive to the United States, as far as method and apparatus claims, the equivalency test would be hard to be seen in Europe.
The same holds for Japan, that is, there is no parallel test. Requiring cooperative working between software and hardware resources, however, may sometimes result in transformation of hardware to a different state. Basically the transformation test would be adopted when an invention does not recite a “particular” machine. In a way, this test supplements the invention including only non-particular machines with particularity in a different way than the machine test. Therefore, as a whole, the Bilski test is more restrictive than the European test, and is virtually similar to Japan. In addition, it seems to be somewhat flexible compared to Japan.
Then, which approaches are more appropriate as far as computer software inventions are concerned? Regarding the machine test, it seems to be ridiculous that putting meaningless limits on the claim scope can transfer an abstract invention into a concrete invention without altering the claim scope. To be a concrete invention, it is not surprising that the certain elements are required in a claim that can draw the line between what is and is not the claimed invention. Consequently, the U.S. and Japanese approaches seem to be more favorable than the European. On the other hand, the wording “meaningful” and “significant” is now debatable in the United States. Underlying reasons to add these words to the Bilski test seem to be analogous to the concept of such terms: “cooperative working,” “concretely realized by using hardware resources,” and “go beyond the normal physical interactions” in Europe and Japan. To ensure stability and predictability, therefore, the language “cooperative working,” “concretely realized by using hardware resources,” and “go beyond the normal physical interactions” may be helpful to construe the terms “meaningful” and “significant”.
How about the transformation test? No analogous test in Europe and Japan means both jurisdictions require some sort of machines, such as technical means and hardware resources. In the case of computer software, software is premised to use with hardware, such as being loaded in and run on a computer, and it inevitably provides some sort of transformation of hardware when it is run on the computer. Therefore, it may be enough to require only the machine test, and in fact, it would be troublesome to draw the line between the transformation, which provides patent eligibility, and the transformation which does not. Moreover, the language “article” might be defined as an electron, molecule or even “signal.” In such a situation, the transformation test would make no sense. Considering this Internet era, requesting too many technical means or hardware resources would not be fit to present business models. There is not doubt that more and more computer software will lose touch with hardware. Therefore, adhering to the machine test might cause a chilling effect on new computer technology and industry. Even though it may be necessary to reconsider and redefine the transformation test, in the near future, the test which focuses on results and/or effects on physical entities would be essential to decide whether it is an abstract idea or not.
As for business methods, if the business methods are realized by using computer software, the problem about abstractness would be identical to other computer software inventions. In addition, business methods have other issues; most business method inventions merely computerize the method that has already been conducted without using a computer, and pure business methods are usually not technical. Unlike the United States, in Europe and Japan, the Article 52(2) EPC and the 2000 guidelines in Japan state that business methods as such are not patentable subject matter, because that lack technical aspects or do not utilize a law of nature. On the other hand, in the United States, business methods as such could be patented as far as they satisfy patent eligibility and patentability inquiries. In the past, therefore, business method inventions, which may be rejected by the EPO and JPO due to lack of patent eligibility, could be patented by the USPTO. Bilski, however, completely changed this situation.
As seen in recent Board’s decisions, it is as if business method inventions were never granted a patent, because the inventions do not include particular machines (without tying machines) and purely business elements are not an “article” (they have no transformation). On the other hand, in Europe, a purely non-technical purpose could be patentable subject matter by simply reciting technical means (Hitachi). Since even a general purpose computer can be technical means, business method inventions will likely be found patent eligible by the EPO, rather than by the USPTO. However, whether such inventions would be patented by the EPO is another story. In the case of Japan, a business method invention could be statutory, when cooperative working between software and hardware resources is observed. Therefore, in most cases, the U.S. approach is now the most restrictive, and the Japanese approach is the second most restrictive.
Thus, although the Federal Circuit once explicitly denied the business method exception, the Bilski test implies that such an exception does exist. There should be business method inventions that are suitable for patent protection however, and at least these inventions have contributed to “promot[ing] the progress of … useful arts.” In fact, the EPO and the JPO, in which an invention directed to subject matter that is nothing more than methods for doing business as such cannot be statutory, have both granted patents for inventions that include business factors in varying degrees. Nevertheless, Bilski seems to eliminate not only purely business methods, but also almost all kinds of business methods from patentable subject matter, even though it may be difficult to draw the line between purely and non-purely business method inventions.
The majority in Bilski did not overrule State Street Bank and AT&T, therefore what kind of business method inventions should be patented remains uncertain. Seemingly, it would be questionable whether a method for use in a telecommunications system as in AT&T could withstand the machine-or-transformation test, because it hardly recites particular machines and a PCI indicator, which may be an article, would no be found transformed into a different state. Thus, setting forth the Bilski test without adjusting the apparent conflict with CAFC precedent might be one of the reasons to cause confusion about what type of business method inventions should be protected.
In Europe and Japan, due to lack of technicality and the utilization of laws of nature, there is some excluded subject matter. For example, mere mathematical methods, mental activities, arbitrary arrangements (methods for playing games), methods doing business and presentations of information would be excluded. The majority in Bilski used the term “mental process” as if the mental process doctrine was revived. As seen in European and Japanese practices, the mental process could be a helpful standard in the United States to assess whether it is purely a business method. Therefore, reintroducing the mental process doctrine might be a more possible way than struggling with troublesome interpretation of the term “article,” although adjustment with reasons for abandonment of the doctrine are imperative.
- V. PATENTABILITY PRACTICES OF EACH COUNTRY
Even after passing the patent eligibility inquiry, an invention has to overcome the novelty and non-obviousness tests to obtain a time-limited exclusive right. In the United States, the novelty and non-obviousness requirements are codified in 35 U.S.C. §§ 102 and 103 respectively. In the case of Europe and Japan, the novelty and inventive step requirements are provided in Article 54 and 56 in the EPC, and Article 29(1) and (2) in the Japan Patent Act. One of the reasons for opponents of patent protection for computer software inventions is that, unlike other technology areas, “software development is very incremental and typically any novelty is based on a mixture of original and derived elements, rather than a single inventive step.” Since, in almost all cases, the most controversial point is non-obviousness in the United States and inventive step in Europe and Japan, the question of how far a computer software invention should be improved so as to obtain patent protection is a fatal issue.
In the United States, regardless of the technological field, the “teaching, suggestion, or motivation test” (TSM test) had been a dominant test employed by the CAFC and the USPTO to decide whether an invention is obvious over prior art until KSR was announced by the Supreme Court in 2007. Before KSR, the examiner had to show a teaching, suggestion or motivation to combine more than two prior art preferences to prove a patent claim was obvious. The teaching, suggestion or motivation had to be found in the prior art, the nature of the problem, or knowledge of a person having ordinary skilled in the art. The TSM test itself was not an inadequate test for obviousness, but rigid application could be inappropriate, such as being “confined by a formalistic conception of the words “teaching,” “suggestion,” and “motivation,” or by overemphasis on the importance of published articles and the explicit content of issued patents.” Thus, since rigid application of the TSM test would be a heavy burden of proof for the examiner to overcome, generally it was relatively easy for applicants to obtain a patent. After KSR, although the Supreme Court did not completely deny the TSM test, it emphasized that it was not the only rationale for § 103 by reaffirming the obviousness test established in Graham v. Deere. The Court suggested seven rationales including the TSM test, combining prior art elements according to known methods to yield predictable results. These rationales basically focus on “predictability.”
Computer software development is usually very incremental and a business method invention is the existing business model or method realized by implementing a computer. Therefore, unlike other technological fields, such as chemistry and biotechnology, computer software inventions would be highly predictable and then would likely be rejected under § 103. At least before KSR the quality of computer software inventions had been one of the serious problems in the United States Considering the nature of computer software inventions, the rigid TSM test would impose too a heavy burden on the examiner, which would not be recognized in Europe and Japan. This would be one of the reasons to generate questionable patents in the United States. Therefore, from the aspect of not only quality, but also international harmonization, the post-KSR approach could be one of the powerful means to resolve that problem. Because of high predictability of computer technology, however, there is a risk of arbitrary examination based on KSR. The USPTO and the courts should observe close attention to that risk.
In Europe, the EPO has adopted the Problem and Solution Approach (PSA) to determine whether a claimed invention has an inventive step under Article 56 EPC. PSA is constituted in three steps. The first step is identifying the closest prior art (CPA), which often possesses the greatest number of features in common with the invention. The second step is establishing the objective technical problem to be solved based on the CPA , and the last step is considering whether the claimed invention, starting from the CPA and the objective technical problem, would have been obvious to a skilled person. In the second step, the examiner identifies the differences, in terms of the claimed technical features, between the claimed invention and the CPA. Thus, under the EPO practice, only technical features and aspects are taken into account to estimate an inventive step, and basically non-technical features are completely disregarded.
Considering a computer-implemented invention case, if the claimed invention does not include a technical character, then it will be rejected under Article 52(2) and (3). If there is a mix of clearly technical and non-technical aspects, the examiner will apply PSA only to the technical aspects, and identify the CPA based solely upon the technical implementation set forth in the claims and description. Next, in the second step, when the differences are not technical, the invention will be rejected under Article 56. If they are purely technical, the invention will be evaluated under the classic PSA approach. If the differences are a mix of technical and non-technical features, the examiner will formulate the objective technical problem, taking into account the requirement specification. When the solution of the technical problem is obvious, the invention will be rejected under Article 56 EPC.
In the case of mix inventions, if the only technical aspect is a conventional computer, the invention cannot be a patent, even if a non-technical aspect would be considered new and non-obvious. Examination would be required only in relating technical aspects, such as the conventional computer. Thus, the invention might be rejected without substantial examination related to the core of the invention. This is a huge difference from the United States and Japan. In those countries, the examination is done for the whole claimed elements regardless of the technical or non-technical elements. As a result, although the EPO notes that it considers a non-obvious technical effect achieved by a non-technical feature in combination with a technical feature, the inventive step of computer-implemented inventions might be difficult to be detected under the PSA approach, as compared to the United States and Japan. In other words, no matter how narrow the EPO restricts the ambit of excluded subject matter (non-technical), the EPO could maintain the quality of patents because non-technical aspects might be nearly disregarded at the inventive-step inquiry.
In the case of Japan, unlike with patent eligibility, the obviousness inquiry for computer software inventions is not different from other technical fields, and the 2000 guideline (a general guideline) not the CS guideline is applied. The basic inquiry set forth in the 2000 guideline is almost identical to the Graham test in the United States; first, finding a claimed invention and one or more cited inventions and selecting one cited invention that is most suitable for reasoning; second, comparing the claimed invention with a cited invention and clarifying the identicalness and the difference; and third, reasoning for lacking an inventive step of the claimed invention on the basis of the contents of the selected invention, other cited inventions (including well-known or commonly used art) and the common general knowledge. As examples of reasoning, the 2000 guideline specifies that selection of an optimal material, workshop modification of design, and mere juxtaposition of features are regarded as obvious. Further, it addresses that the close relationship of technical fields, close similarity of problem to be solved, close similarity of function, work or operation and suggestion shown in the contents of cited inventions are possible causes or motivation to combine cited inventions. Meanwhile, if an advantageous effect compared to cited inventions can be clearly identified from descriptions in the specification and the drawings, it is taken into consideration as a fact to support an affirmative interference of its inventive step. Thus, although the expression may be different, the Japanese approach in the 2000 guideline is remarkably analogous to the Supreme Court decision in KSR.
The CS guideline, however, adds some explanation relating to its reasoning: combining technologies used in different fields and applying them to another field is usually to be within the exercise of an ordinary creative activity of a person skilled in the art as long as there is no technical difficulty for their combination or application. The guideline reasons that combining a method or a means used in different fields or applying them to another field in order to achieve an intended objective are usual attempts in the field of software technology. Thus, in the case of computer software-related inventions, non-obviousness is defined in the negative. In addition, the guideline provides examples of exercising ordinary creative activity expected of a person having ordinary skill in the art, such as an application to other fields, implementation by software of functions that are otherwise performed by hardware, systematization of human transactions or reproduction of a known event in computerized virtual space. Thus, if a known business method in one field is applied in another field, such application would not be regarded as an inventive step. In addition, the effects of the invention, such as “can be processed quickly,” “can process a large amount of data” and “can be uniform results” cannot usually be considered an unforeseeable result and an inventive step, since these effects are often obtained as a result of computerization.
After KSR and Bilski, computer software inventions seem to have entered a difficult time for being granted a patent in the United States. However, the standard created by the court does not go far over the edge, rather it is coming closer to other countries’ approaches. Therefore, the result for the quality of the patents issued after these two cases is also expected to be closer, but should be carefully observed. The United States approach demonstrates great similarities to that of Japan, but it is remarkably different from the EPO approach, especially by virtually ignoring non-technical aspects. In the United States, since usefulness is the core concept of the patent system, disregarding the useful non-technical effects might be considered to go against the United States Constitution and the United States patent law. Also, in Japan, computer programs (non-technical) can be assumed to be potentially utilizing a law of nature, therefore the EPO standard should not be accepted. In this regard, the EPO may show a more restrictive attitude toward computer software inventions than the United States and Japan. In other words, the matter of patent eligibility may not be so serious in Europe compared to the other two countries as far as the EPO adopts its Problem and Solution Approach.
- VI. CONCLUSION
With the development of computer software and information technologies, one of the significant issues in the future will be whether requiring physical linkage between software and hardware can catch up with the tide of such technologies. Now, the United States, the EPC member countries and Japan all rely heavily on tying machine requirement to draw the line between what is and is not statutory subject matter. However, computer software inventions are more and more losing touch with hardware. In this situation, the underlying concept of the transformation test seems to be important, and not only the United States, but also other countries should consider that test seriously for the near future.