Pharmascience is not a software case. Its method of reading the Patent Act may nevertheless provide the missing Canadian bridge to the UK and EPO approach.
In two recent posts, I argued that Canada’s software patent problem is structural. Subject-matter analysis has been asked to do work that properly belongs to novelty and obviousness. The result is an unstable threshold inquiry into whether an invention feels sufficiently “technical,” “concrete,” or inventive — often without the discipline of prior art or a proper evidentiary record.
In February, the UK Supreme Court addressed essentially that problem in Emotional Perception AI Ltd v Comptroller General of Patents. It rejected the Aerotel “actual contribution” analysis because asking what an inventor has “really added to human knowledge” improperly blends the threshold question of whether there is an invention with the separate requirements of novelty and inventive step. The Court adopted a low software-specific eligibility threshold, followed by a filtering step and a proper inventive-step analysis.
Now the Supreme Court of Canada may have supplied the Canadian half of that argument.
A Pharmaceutical Case With a Broader Structural Point
On July 17, 2026, the Supreme Court released Pharmascience Inc. v. Janssen Inc. The case concerned dosing regimens and the prohibition against patenting methods of medical treatment. It did not concern software, artificial intelligence, computer programs, or abstract ideas.
The majority held that methods of medical treatment remain unpatentable where a claim seeks to monopolize professional medical skill and judgment. It nevertheless upheld Janssen’s patent because the claimed dosing regimens did not require that skill and judgment once the physician had selected the regimen.
One may agree or disagree with placing the professional-skill boundary within subject matter. I am generally comfortable with there being some boundary around the patenting of professional judgment, although enablement, claim clarity and overbreadth may often be more precise tools for enforcing it.
But the broader importance of Pharmascience lies not in the particular boundary the Court drew. It lies in how the Court interpreted section 2.
Keeping the Statutory Questions in Their Own Lanes
Section 2 of the Patent Act defines an invention as: “any new and useful art, process, machine, manufacture or composition of matter,” or an improvement in one of those things.
That drafting creates an obvious temptation. Because “new” and “useful” sit directly beside the enumerated heads, they can be allowed to bleed into the interpretation of “art,” “process,” “machine,” “manufacture” and “composition of matter.”
Once that happens, subject matter becomes a preliminary novelty or utility analysis. A decision-maker asks whether the claimed thing is really new, really useful, really technical, or really inventive enough to count as an art or process. Novelty and utility are then assessed again under their ordinary doctrines, with obviousness separately assessed under section 28.3 of the Patent Act.
That is duplication, but worse, it is duplication using different tools. At the subject-matter stage, the analysis tends to be intuitive and categorical. At the novelty, obviousness and utility stages, it is supposed to be structured, evidence-based and tied to defined legal standards.
In Pharmascience, the Supreme Court kept those questions separate. The majority described patentable subject matter as a necessary but insufficient condition for validity, then separately identified novelty, non-obviousness and utility as additional requirements. It expressly noted that none of those requirements was at issue. The sole issue was patentable subject matter.
The Court then stayed within the enumerated-heads analysis. The question was whether the claimed subject matter constituted an “art” or “process,” as those terms have been judicially interpreted, or instead claimed professional medical skill and judgment.
The majority also rejected the concurrence’s proposal to move much of the professional-skill problem into utility. It did so partly because the proposed concepts of operability, reproducibility and control would alter established utility doctrine, under which the invention need only be capable of a practical purpose or actual result and satisfy the low threshold of a scintilla of utility.
In other words, the majority did not permit utility to expand merely because another doctrine was difficult to apply. The majority did not let “useful” do hidden work inside section 2, as was being suggested by the concurrence. It kept “machine,” “art,” and “process” in their own lane – cleanly distinct from the terms “new and useful”.
Harvard Mouse, Now Pharmascience
This is consistent with the Supreme Court’s approach in Harvard College v. Canada (Commissioner of Patents).
There, the Court did not ask whether the oncomouse was sufficiently new, useful, inventive or technologically meritorious to deserve a patent. It asked whether a higher life form was a “manufacture” or “composition of matter” within section 2.
In Pharmascience, the Court similarly asked whether the subject matter fell within “art” or “process,” properly understood, or crossed the judicial boundary surrounding professional medical skill and judgment.
One may dispute the Court’s definitions of “manufacture,” “composition of matter,” “art” or “process.” But the method is consistent: construe the claim, identify the statutory head, and determine whether the subject matter falls within that head as judicially interpreted.
But what about Shell Oil?
Shell Oil involved a newly discovered use for old compounds, and the Court treated the application of that new knowledge to a practical result as a “new and useful art.” In other words, it was a statutory subject matter case that embedded “new and useful” directly into the question of what was statutory subject matter. Therefore, the Court necessarily had to consider the terms “new and useful”; but we can note that Harvard Mouse was about a mouse, Pharmascience is about a method of medical treatment, and software is, well, about software — the terms “new and useful” are not embedded into the technology itself. Read in context, Shell Oil is better understood as a new-use/practical-application case, not a pure enumerated-head statutory subject matter case, and not as authority for allowing “new and useful” to spill over and redefine the statutory heads in every subject-matter dispute. The point was that the invention lay in the discovered use and its practical embodiment; Shell Oil does not undermine the basic discipline of keeping subject matter and merits separate.
The Connection to Emotional Perception
That methodology closely matches the structural reasoning of the UK Supreme Court in Emotional Perception.
The UK Court held that whether the claim defines an “invention” must be addressed first and separately from novelty and inventive step. It rejected Aerotel because its “actual contribution” test jumped immediately to what the inventor had added to the known art. That was not really an eligibility analysis. It was an impressionistic merits analysis conducted in the wrong doctrinal place.
For computer-implemented inventions, the UK Court therefore adopted a low threshold: the use of technical means or hardware is enough to clear the initial eligibility hurdle. The harder analysis occurs later. Features that do not contribute to the technical character of the invention are filtered, and the surviving features are assessed against the prior art under novelty and inventive step.
Canada need not adopt every feature of EPO practice to adopt the same statutory discipline. Indeed, the UK Supreme Court retained the domestic Pozzoli approach to inventive step rather than requiring wholesale adoption of the EPO problem-and-solution method. When it comes to software patents, Canada can work within its own jurisprudence and still align with Emotional Perception and EPO Practice.
Where This Leaves Canadian Software Patents
To be clear, Pharmascience says nothing directly about software. Its value here is repeated structural direction from the Supreme Court of Canada as to how to consider an enumerated head under Section 2. Together with Harvard College, and distinguishable from Shell Oil, it confirms that section 2 is about the statutory kind of thing claimed — art, process, machine, manufacture, composition of matter, or improvement — as those terms have been judicially interpreted.
That matters because Schlumberger, later Federal Court cases, and CIPO practice all sit below the Supreme Court of Canada. They have to fit the Court’s section 2 framework, not stretch it.
A software analysis that looks for an “actual invention” or “actual contribution,” treats the computer as merely conventional, and then asks whether what remains is sufficiently technical risks repeating the same mistake rejected in Emotional Perception: turning the threshold inquiry into disguised novelty and inventive-step analysis.
For software, the cleaner path is straightforward, as argued in my earlier posts CIPO’s New Subject-Matter Guidance — and the Structural Problem It Can’t Fix and The UK Supreme Court Just Fixed Software Patent Eligibility — Canada Should Follow.
To reiterate: first, construe the claim and ask whether it defines a statutory art, process, machine, or improvement, rather than a mere scientific principle, abstract theorem, or disembodied idea. In the case of software, we can acknowledge both the UK Supreme Court and the EPO, that a low threshold can be applied to software related inventions. Second, assess the invention on the merits under obviousness. The point of keeping those questions separate is to stop novelty and obviousness from being smuggled into section 2.
The next Canadian software case should take the architecture from Pharmascience, Harvard Mouse and Emotional Perception, seriously.
PCK Intellectual Property is a leading North American intellectual property law firm based in Canada, specializing in the preparation, prosecution, registration, enforcement, and strategic management of patent and trademark portfolios for global markets. Our team of seasoned patent agents, trademark agents, engineers, scientists, and IP lawyers supports clients across diverse technologies, with expertise in trademark law, trademark registration in Canada, trademark applications, trademark prosecution, trademark opposition proceedings, trademark portfolio management, trademark protection and enforcement, brand protection strategy, coordinating cross-border IP strategies including PCT national phase filings, Madrid Protocol trademark registration, and international patent filing in Canada. Whether you are a startup seeking IP strategy guidance, an inventor exploring how to protect an idea in Canada, or a US or foreign company looking to expand IP protection to Canada, contact us today to explore how we can help secure and manage your intellectual property.
The contents of this article are provided for general information purposes only and do not constitute legal or other professional advice of any kind.