The landscape of patentable subject matter for software inventions is complex and continues to evolve. US patent law includes judicially created exceptions to patentable subject matter like laws of nature, natural phenomena, and abstract ideas. This presents a real challenge for software inventors working with algorithms, which oftentimes resemble mathematical logic, creating a grey area between patentable inventions and unpatentable ideas. This ambiguity becomes more pronounced under the contemporary Mayo-Alice test, which sets a high bar for software patent eligibility. In this blog, we will explore the current subject matter challenges in software patenting, review the key tests used by courts, and look at potential reforms that might help clarify the issue.
Back Then: Machine-or-Transformation Test
Patenting software used to be simpler under the machine-or-transformation test. To be patentable, an invention had to be tied to a specific machine or transform something into a different state or thing. For software, this was straightforward since computer programs typically rely on machines to run and often transform data into new forms.
In Bilski v. Kappos (2010), algorithms were considered patentable because they represented specific steps to operate a computer. Although algorithms are based on abstract mathematical concepts, they were seen as distinct from human thought. But if the software simply mirrored an abstract idea without adding anything new, it wasn’t patentable. This was illustrated in Gottschalk v. Benson (1972), where a method of converting signals from binary coded decimal form to pure binary form was deemed unpatentable because it didn’t go beyond a basic mathematical idea.
Today: Mayo-Alice Test
While the older test was more lenient on software patent eligibility, it wasn’t precise enough to deal with modern software inventions. The court introduced the Mayo and Alice rulings, creating a stricter two-step test for determining patent eligibility. Here’s how it works:
1. Step One: Does the software claim involve a patent-ineligible concept, including law of nature, natural phenomenon, or abstract idea?
2. Step Two: If so, does the claim contain an inventive concept sufficient to transform it into a patent-eligible application?
In Mayo Collaborative Services v. Prometheus Labs, Inc. (2012), the Court emphasized that a patent claim must involve something more than just “well-understood, routine, or conventional” steps. Then in Alice Corp. v. CLS Bank Int’l (2014), the Court ruled that simply applying an abstract idea on a computer isn’t enough for a patent claim.
For software inventors, this means that to pass the Mayo-Alice test, they need to show that their software does more than automating an existing idea. It must improve computer functionality or solve a specific technological problem—something more concrete than just running a process on a computer.
Why It’s So Confusing and How We Might Fix It?
While the Mayo-Alice test offers some guidance for distinguishing between patent-eligible software inventions and abstract ideas, the boundaries between the two remain unclear, especially when it comes to deciding what counts as an “inventive concept.” This lack of clarity makes it difficult to predict whether a software invention qualifies for a patent and leaves a lot of room for interpretation by patent examiners and courts
The restrictive nature of the Mayo-Alice framework can discourage innovation. Many inventors struggle to frame their inventions in a way that meets the strict criteria, which can prevent legitimate technological advancements from being patented.
How can this be resolved? Ideas have been raised to alleviate the present uncertainty.
One option is to relax the patentable subject matter standard and defer the above issues to subsequent review on novelty and non-obviousness. By framing unpatentable elements as prior art, examiners can exclude them from 35 U.S.C. §§ 102 and 103 evaluations, and focus on identifying novelty and non-obviousness in other parts of the claim. However, this approach could cause a tragedy of anti-commons problem, where too many patents create excessive litigation and hinder technological progress. Plus, it might just shift the issue to a later stage in the process.
Another proposal is to create a categorical, codified list of excluded subject matter. The Patent Eligibility Restoration Act of 2023 (PERA), proposed by Senator Thom Tillis, aims to add express exceptions to patent eligibility under § 101, such as mathematical formulas and mental processes.
By providing statutory guidelines, this approach aims to enhance predictability for inventors. However, it’s important to recognize that this bill is not a catch-all solution to eligibility litigation. While it may alleviate some ambiguity in the Mayo-Alice framework, it introduces new challenges about how the revised § 101 will be applied. For instance, how will courts interpret what it means for a process to be “substantially economic, financial, business, social, cultural, or artistic”? And how can we determine if a computer is truly essential to the process or if the task could be performed without one? These newly-introduced uncertainties might transform the § 101 analysis into a fact-heavy inquiry, requiring more extensive discovery than courts typically allow at this stage.
Conclusion
The patentability of software inventions continues to be a challenging issue for inventors, patent examiners, and courts. While the Mayo-Alice framework has introduced some rigor to prevent the patenting of mere abstract ideas, its lack of clarity creates hurdles and confusion for software patent applications.As the debate over patentable subject matter evolves, potential reforms, such as relaxing the current standards or codifying clearer exceptions, may offer hope for a more predictable and innovation-friendly patent system. In the meantime, patent applicants must carefully navigate the nuances of the Mayo-Alice test to frame their software inventions in ways that demonstrate genuine improvements.