Bilski v. Kappos, Mayo v. Prometheus, and CLS Bank v. Alice Corp.: The Death of the Machine-or-Transformation Test and the Effects on Software Patent Prosecution

By   /   May 11, 2013 


Bilski v. Kappos, Mayo v. Prometheus, and CLS Bank v. Alice Corp.:
The Death of the Machine-or-Transformation Test and the Effects on Software Patent Prosecution

Jan Lucas


I           Introduction 

The status of computer software as patent-eligible subject matter has taken many turns over the years.  When computers were first introduced, there was no doubt that the hardware components of these new machines were of patent-eligible subject matter.  The same, however, cannot be said of the software that came with it.  During the early stages, questions focused on whether software should be patent eligible at all or whether software should be protected only by copyright.1  Over the subsequent decades, it has become more or less accepted2 that at least some form of patent protection should be afforded to software, but the scope of that protection is by no means settled law.

With its decision in Mayo v. Prometheus,3 the Supreme Court once again left the software industry with confusion as to the trajectory of the fight over this issue.  In Mayo, a unanimous Court, in an opinion written by Justice Breyer, held that patents4 which claimed methods for calibrating the proper dosage of drugs for treating particular diseases were invalid for effectively claiming the underlying laws of nature.5  As early as 1972, the Supreme Court held that such “[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.”6  Thus, this holding in Mayo did not lay down any new law.  In fact, the Court merely reaffirmed a principle that it has consistently applied for over forty years.  The unsettling part of the decision, however, was the breadth with which the Court applied this long established principle.


II         The Supreme Court’s Benson-Flook-Diehr Trilogy of Decisions

In Gottschalk v. Benson7 the Court held that a mathematical algorithm8 contained within a computer program, without substantial practical application, is not patentable.9  The patent application at issue in the case “claimed a method for converting binary-coded decimal (BCD) numerals into pure binary numerals.”10  Without limitation, the claims sought to cover any use of the conversion method in any general-purpose computer.11  The Court found this to be a claim to a general mathematical algorithm and not to any practical application of that algorithm, which rendered it unpatentable.12

Mathematical algorithms, like natural phenomena, are abstract ideas, which are inherently unpatentable.13  However, a practical application of some useful structure created with the aid of a mathematical algorithm may be patentable.14  The distinction the Court drew in Benson was that using a mathematical algorithm in a computer program does not constitute a practical application if the program itself is also generic and does not perform a specific function or create a “useful, concrete, and tangible result”15 other than the execution of the mathematical algorithm.16  Thus, while limiting the scope of software patents, the Benson decision, and its progeny, did exactly what the patent law was designed to do—grant limited monopolies to reward inventors for their discoveries, while limiting the obstacles on further innovation and development in the art.

In Parker v. Flook17 the Court again rejected a patent application claiming the use of a mathematical algorithm, though this time the process was limited to a particular field.18  The patent application in question claimed a process for calculating and updating the alarm limit while monitoring conditions during the catalytic conversion process in the petrochemical and oil-refining industries.19  The only novel feature of the claimed invention was the use of a mathematical algorithm.20

While the invention in Flook limited its claim to the algorithm to the petrochemical and oil-refining industries, the Court nevertheless rejected the claims based on a finding that, other than the use of the mathematical algorithm to calculate the alarm limits, the invention did not contain anything new in the art.21  Simply adding “post[-]solution activity,” activities performed after the mathematical algorithm is solved, does not “transform an unpatentable principle into a patentable process.”22

Thus, the outcome of the Flook decision resulted in further limiting the scope of what can be claimed in a software patent.  Implementing a mathematical algorithm in a process that is already known in the art and simply limiting it to a particular field or environment will not circumvent the prohibition against patenting abstract ideas.23  Similarly, adding “insignificant post[-]solution activity” to a process that uses a mathematical algorithm to perform a function, but does not contain any new invention beyond the use of the algorithm, also does not circumvent the prohibition.24

However, in Diamond v. Diehr25 the Supreme Court held that “an application of a law of nature or mathematical [algorithm] to a known structure or process may well be deserving of patent protection.”26  This was a partial reversal of course from the Flook decision and certainly narrowed the application of the Benson and Flook decisions to software patents.  The patent application at issue in Diehr claimed a new method for “molding raw, uncured synthetic rubber into cured precision products” by means of a mathematical algorithm27 to complete certain steps through the use of a computer.28  The Court emphasized that analysis of such claims should focus on the invention as a whole and not “dissect the claims into old and new elements and then . . . ignore the presence of the old elements in the analysis.”29  New combinations of steps in process claims may still be patentable even if none of the constituent parts of the combination is new in the art.30

The Court distinguished Diehr from Flook by noting that “when a claim containing a mathematical [algorithm] implements or applies that [algorithm] in a structure or process which, when considered as a whole, is performing a function which the patent laws were designed to protect,” it may be patentable.31  The Court found that the claim in Diehr was geared towards patent-eligible subject matter because it transformed matter from one state to another, whereas the claim in Flook was unpatentable because it merely performed numerical calculations.32  While this was a very narrow distinction, and some might argue that there was no difference at all, the Court did not overrule Flook, opting instead to limit its holding.33  Mathematical algorithms remained unpatentable, however, because the claim in Diehr was not drawn to a mathematical algorithm but rather to an industrial rubber molding process that simply contained a mathematical algorithm as one of its steps, the Court found the patent application’s claims to be of patent-eligible subject matter.34


III        Subsequent Developments

Having received guidance from the Supreme Court in the Benson-Flook-Diehr line of decisions, the lower courts were left to determine what was a useful, concrete, and tangible result sufficient to satisfy this requirement.35  Over the years, in cases such as In re Alappat,36 State Street Bank & Trust v. Signature Financial Group,37 and AT&T Corp. v. Excel Communications,38 the Federal Circuit ruled frequently on what types of inventions met the requirements, but it was not until more recently that the Federal Circuit set out to adopt a particular test to answer that question.  In a decision from 2008,39 the Federal Circuit held that the “machine-or-transformation test”40 should thenceforth be used as the sole test for determining whether an invention using laws of nature is sufficiently concrete to satisfy the requirements of § 101,41 only to have that holding reversed by the Supreme Court.

In Bilski v. Kappos (Bilski II),42 another case involving a mathematical algorithm, the Supreme Court considered whether a patent application claiming a method of hedging risk in the field of commodities trading in the energy market was directed to statutory subject matter as required by § 101.43  In that case, the Court affirmed the rejection of all the claims in the application for trying to claim an abstract idea.44  In addition, the Court held that the machine-or-transformation test is not the sole test for determining whether some claimed inventions are processes under § 101.45

On the question of the patent eligibility of the process under § 101, the court held, “The concept of hedging described in claim 1 and reduced to a mathematical [algorithm] in claim 4, is an unpatentable abstract idea, just like the algorithms in Benson and Flook.”46  The other claims were simply examples of how hedging could be used, and tried to limit the use of the abstract idea embodied in them to a particular field by adding post-solution components, which is exactly what the Court prohibited in Flook.47  As such, the Court held that the claims were all correctly rejected, but it based its reasoning on the Benson-Flook-Diehr line of cases, and not the machine-or-transformation test.48

Regarding the machine-or-transformation test, the Court held that while that test “is a useful and important clue, an investigative tool, for determining whether some claimed inventions are processes under § 101[,] . . . [it] is not the sole test for deciding whether an invention is a patent-eligible process.”49  While that test may be well suited to evaluate Industrial Age processes, such as inventions involving tangible matter, the Court reasoned that it was not equally suitable for inventions from the Information Age, which may not necessarily be in physical or tangible form.50  Specifically, the patentability of inventions involving “software, advanced diagnostic medicine techniques, and inventions based on linear programming, data compression, and the manipulation of digital signals” would be questionable under this test.51  Instead, when courts evaluate patentability of emerging technologies, they should not be limited to asking “questions of such intricacy and refinement that they risk obscuring the larger object of securing patents for valuable inventions without transgressing the public domain.”52


IV        Mayo v. Prometheus and CLS Bank

Following the Bilski II decision, the software industry again breathed a sigh of relief, after having faced a significant limitation on the patentability of software, which was imposed by the machine-or-transformation test.  While that test has not been completely rejected, the Supreme Court’s dicta, suggesting more of a “big picture analysis,” indicate support for a case-by-case and fact specific evaluation of process claims, instead of the more rigid machine-or-transformation test.  However, while holding that the machine-or-transformation test is not the sole test for this issue, the Court did not suggest an alternative test.  So, as unfavorable as the machine-or-transformation test was, the uncertainty of what should be used in its stead turned out to be equally unsettling.

The general consensus was that the Federal Circuit, without any other test to use, would simply continue to apply the machine-or-transformation test.  And it did so, until Mayo.  In Mayo, the Supreme Court reversed the Federal Circuit on that exact point, but not in the way the industry had come to expect.  Prior to Mayo, it was more or less safe to assume that if a process patent satisfied the machine-or-transformation test, it met the requirements of § 101.  But in Mayo, the Court held a patent invalid even though the Federal Circuit had found—twice53—that it satisfied the machine-or-transformation test.  This ruling now reveals the possibility that even the rigorous machine-or-transformation test does not assure compliance with § 101, which leaves the software industry with a lot of new questions.

On May 10, 2013, the Federal Circuit announced its decision in its en banc rehearing of CLS Bank v. Alice Corp.54  In that case, the court affirmed the district court’s holding that Alice Corporation’s patent claims were invalid because they were drawn to an abstract idea.55  The patents56 at issue claimed certain methods of reducing settlement risk in financial transactions by using a third party intermediary with the power to verify that both parties to the transaction can fulfill their obligations prior to allowing the transaction to be completed.57  In the leading plurality opinion,58 in which five judges joined, the court emphasized that patent “claims should not be coextensive with . . . abstract ideas” but should include substantive limitations that “add significantly more to the basic idea, with the result that the claim covers significantly less.”59  “What matters is whether a claim threatens to subsume the full scope of a fundamental concept, and when those concerns arise, we must look for meaningful limitations that prevent the claim as a whole from covering the concept’s every practical application.”60

While the court, in granting the rehearing, had asked for briefing on what new test should be adopted to replace the machine-or-transformation test, the court did not adopt such a rigid new test in its decision.61  Some guidelines were given, but those were largely just restatements of previously announced principles and Supreme Court precedent dealing with § 101.  The court simply emphasized again that the invention must fall within one of the four statutory categories set out in § 101 and not one of the judicially crafted exceptions–natural laws, natural phenomena, or abstract ideas.62  If the claims fall under one of the exceptions, the inquiry proceeds to the preemption analysis to ensure that the claims are not coextensive with the underlying abstract idea.63  So, while it remains to be seen how this will be interpreted and applied by the Patent Office and the lower courts, it appears that the claim-by-claim approach to determining a patent’s compliance with § 101 will remain, at least for now.


V         Conclusion

The Supreme Court suggested in Bilski II that the machine-or-transformation test should be replaced, but it did not articulate a new test with which to replace it.  As such, the lower courts continued to apply the machine-or-transformation test as the primary test to determine whether claims were drawn to patent-eligible subject matter under § 101.  Bilski II was seen simply to stand for the position that failing to satisfy the test does not render a claim patent-ineligible.

Prior to Mayo, the test, if nothing else, was at least a method for distinguishing between claims that were certainly patent eligible under § 101—those that satisfied the test—and claims that may not be—those that did not satisfy the test.  This changed in Mayo, when for the first time the Court held invalid a patent that satisfied the machine-or-transformation test.  With this decision, the test became, essentially, useless.  Following Bilski II the test no longer served as an indicator of patent ineligibility if a claim failed the test, and, following Mayo, not as an indicator of patent-eligibility if it satisfied the test.  And, while the Federal Circuit appeared poised to adopt a new test in CLS Bank, the court did not announce any clear new test.

The Supreme Court’s dicta in Bilski II regarding using a test that is tailored toward Information Age inventions seems favorable to the software industry, but it remains to be seen what test will emerge.  And with the persistent threat that software may someday be ruled unpatentable as a whole, the hope is that a new test will establish once and for all that software is, and should be, patent-eligible subject matter within the meaning of § 101.



facebooktwittergoogle_plusredditpinterestlinkedinmailby feather
facebooktwitterlinkedinrssby feather


  1. Gottschalk v. Benson, 409 U.S. 63, 72 (1972).  While the decision in Benson focused on a particular type of process claim within a software patent, the Court also held that software as a whole was unpatentable—a position it would not officially reverse until 1981.
  2. There are, however, still some holdouts, including Judge Posner of the 7th Circuit, who recently suggested that software should not be eligible for patents at all.  Dan Levine, Judge Who Shelved Apple Trial Says Patent System Out of Sync, Reuters (Jul. 5, 2012 8:42 AM),
  3. Mayo Collaborative Services v. Prometheus Laboratories, Inc., 132 S. Ct. 1289 (2012).
  4. U.S. Patent No. 6,355,623 and U.S. Patent No. 6,680,302.
  5. Mayo, 132 S. Ct. at 1305.  While this case dealt with pharmaceutical patents, the effects of the holding reached far beyond just the pharmaceutical industry, and certainly into the software industry, because of the point of law on which it was decided.  The “laws of nature” exception to patentability holds that any patent seeking to claim a basic law of nature or other natural phenomenon—such as, i.e., Newton’s Law of Universal Gravitation—is, on its face, invalid.  This also applies to mathematical algorithms, and it is in this sphere that the Court’s opinion has its effect on software.
  6. Benson, 409 U.S. at 67.
  7. 409 U.S. 63 (1972).
  8. Mathematical algorithms are the most common laws of nature to apply in the software industry.  Mathematical algorithm, mathematical formula, mathematical principle, and mathematical equation are all used interchangeably by the courts in discussing this concept, even though the scope of the term has never been fully defined, and each of those terms are subject to a variety of definitions, some not always synonymous.  For clarity, this Article only uses the term “mathematical algorithm” to refer to this concept.
  9. Benson, 490 U.S. at 73.
  10. Id. at 64.
  11. Id.
  12. Id. at 65.
  13. Id. at 67; O’Reilly v. Morse, 56 U.S. 62 (1853).
  14. Benson, 409 U.S. at 67 (quoting Mackay Co. v. Radio Corp., 306 U.S. 86, 94 (1939)).
  15. In re Alappat, 33 F.3d 1526, 1544 (Fed. Cir. 1994).
  16. Id. at 67-68.
  17. 437 U.S. 584 (1978).
  18. Id. at 585.
  19. Id. at 585-86.
  20. Id. at 585.
  21. Id. at 594.
  22. Flook, 437 U.S. at 590.
  23. Diamond v. Diehr, 450 U.S. 175, 191 (1981).
  24. Id. at 191-92.
  25. 450 U.S. 175 (1981).
  26. Id. at 187.
  27. The algorithm used in this case was the Arrhenius equation.
  28. Diehr, 450 U.S. at 177.
  29. Id. at 188.
  30. Id.
  31. Id. at 192.
  32. Id. at 186.
  33. The Court has as recently as 2010 still cited Flook as good law.  Bilski v. Kappos, 130 S. Ct. 3218, 3230 (2010).
  34. Diehr, 450 U.S. at 192-93.
  35. Alappat, 33 F.3d at 1544.
  36. Id. (upholding a patent in which “the claimed invention as a whole was directed toward forming a specific machine that produced the useful, concrete, and tangible result of a smooth waveform display”).
  37. 149 F.3d 1368 (Fed. Cir. 1998) (holding “that the transformation of data, representing discrete dollar amounts, by a machine through a series of mathematical calculations into a final share price, constitutes a practical application of a mathematical algorithm, formula, or calculation, because it produces ‘a useful, concrete, and tangible result .   . .’ ”).
  38. 172 F.3d 1352 (Fed. Cir. 1999) (upholding a patent claiming a method for calculating charges for long distance telephone calls which uses multiple different operators’ networks by means of Boolean algebra calculations).
  39. In re Bilski, 545 F.3d 943 (Fed. Cir. 2008) (Bilski I).
  40. Under this test a “claimed process is surely patent-eligible under § 101 if: (1) it is tied to a particular machine or apparatus, or (2) it transforms a particular article into a different state or thing.”  Bilski, 545 F.3d at 954.  To put it differently, the process must be implemented with a particular machine—one specifically devised and adapted to carry out the process in a way that is not concededly conventional and is not trivial.  Or, in the alternative, the process transforms an article from one thing or state to another.
  41. 35 U.S.C. § 101 (1952).  “Whoever 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 therefor, subject to the conditions and requirements of this title.”  Id.  Unless otherwise noted, all statutory references cited in this Article are codified under Title 35.
  42. 130 S. Ct. 3218 (2010).
  43. Id. at 3223.
  44. Id. at 3231.
  45. Id. at 3227.
  46. Id. at 3231.
  47. Bilski II, 130 S. Ct. at 3231.
  48. Id.
  49. Id. at 3227.
  50. Id.
  51. Id.
  52. Bilski II, 130 S. Ct. at 3227.
  53. The final decision from the Supreme Court in Mayo was the second time the Court granted certiorari in that case.  The Court had previously vacated and remanded (Mayo Collaborative Services v. Prometheus Laboratories, Inc., 130 S. Ct. 3543 (2010)) the Federal Circuit’s decision (Prometheus Laboratories, Inc. v. Mayo Collaborative Services, 581 F.3d 1336 (Fed. Cir. 2009)) rendered prior to Bilski II, in light of the new development in Bilski II.  On remand, the Federal Circuit reapplied the machine-or-transformation test, and again found the patent valid (Prometheus Laboratories, Inc. v. Mayo Collaborative Services, 628 F.3d 1347 (Fed. Cir. 2010)).  Subsequently, the Supreme Court again granted certiorari, and this opinion followed.
  54. CLS Bank Int’l v. Alice Corp., No. 2011–1301, 2013 WL 1920941 (Fed. Cir. May 10, 2013).
  55. Id. at *1.
  56. U.S. Patent No. 5,970,479, U.S. Patent No. 6,912,510, U.S. Patent No. 7,149,720, and U.S. Patent No. 7,725,375
  57. CLS Bank, 2013 WL 1920941, at *1-2.
  58. A total of seven opinions were filed in the case, including some “Additional Reflections” by Chief Judge Rader.  See id. at *59-61.  The only point on which a majority of the ten judges on the panel (Judge Taranto did not participate in the decision) agreed was the per curiam opinion affirming the judgment of the district court.  Only a plurality of five judges joined the leading opinion, written by Judge Lourie.
  59. Id. at *8.
  60. Id.
  61. The Federal Circuit’s first decision in CLS Bank (CLS Bank Int’l v. Alice Corp., 685 F.3d 1241 (Fed. Cir. 2012)) was handed down on July 9, 2012, in which the federal circuit reversed the district court’s decision (CLS Bank Int’l v. Alice Corp., 768 F. Supp. 2d 221 (D.D.C. 2011)) holding the claims invalid.  That decision was vacated and a rehearing en banc granted on October 9, 2012 (CLS Bank Int’l v. Alice Corp., No. 2011–1301, 2012 WL 4784336 (Fed. Cir. Oct. 9, 2012)).
  62. CLS Bank, 2013 WL 1920941, at *9.  The court also noted that “[a]n ‘inventive concept’ in the § 101 context refers to a genuine human contribution to the claimed subject matter,” though it is unclear how that differs from the “flash of genius” approach which was rejected by Congress in the 1952 revision of the patent statutes.  Id. at *10.
  63. Id.