The More Things Change: Dusome Returns to CIPO
On May 5, 2026, the Patent Appeal Board issued its preliminary redetermination in the patent application at the heart of Dusome v Canada (Attorney General), the Federal Court decision that prompted the Canadian Intellectual Property Office (CIPO) to revise its patentable subject matter framework. In our previous comment, we described CIPO’s March 2026 Practice Notice as “two steps forward,” praising the abandonment of the “actual invention” as a standalone analytical construct while cautioning that the accompanying examples risked perpetuating old habits under new labels. The Dusome redetermination offers the first opportunity to test whether the revised framework makes a practical difference, and the early returns suggest it may not.
Background: The Federal Court’s Guidance
As we previously discussed, the Federal Court in Dusome found that the Commissioner of Patents had erred in refusing the application – not because the claims were necessarily patentable, but because the Commissioner had improperly conducted the subject matter assessment. The Court identified two principal errors: the Commissioner had treated the identification of the “actual invention” as a standalone inquiry separate from purposive construction (the lens through which patent claims are interpreted) and had failed to properly construe the claims before assessing patentability. The Court directed the Commissioner to reconsider the application “afresh” based on the amended proposed claims and in accordance with its reasons.
The application itself relates to a wagering poker game that combines elements of Texas Hold’em and Pai Gow poker. The claims cover both physical-card and computer-implemented versions of the game.
The Redetermination: New Labels, Same Conclusion
The Board’s redetermination follows the purposive construction framework set out by the Supreme Court of Canada in Free World Trust v Électro Santé Inc and Whirlpool Corp v Camco Inc. It identifies the skilled person, canvasses the relevant common general knowledge, construes the meaning of key claim terms, and determines the essential elements of the claims. So far, so good. But it is what comes next that matters: the Board applies analysis drawn from the Federal Court of Appeal’s decision in Schlumberger Canada Ltd v Commissioner of Patents as articulated in the March 2026 Practice Notice, asking whether the claimed invention can be distinguished from Schlumberger and whether it satisfies the physicality requirement set out in Amazon v Canada (Attorney General) and Canada (Attorney General) v Benjamin Moore. In doing so, the Board reaches the same conclusion as before: the claims are directed to non-patentable subject matter.
The reasoning will be familiar. The Board identifies the only “new knowledge” or “discovery” as the set of rules governing the poker game. The physical cards (claims 1-21) and computer components (claims 22-24) are characterized as well-known tools used in well-known ways. The implementation of an abstract set of rules on those well-known instruments, the Board concludes, does not satisfy the physicality requirement without “something more.” The Board also considers Progressive Games and concludes that the modifications to poker in this application do not amount to a new and innovative method of applying skill or knowledge within the meaning of Shell Oil Co v Commissioner of Patents. The bottom line is the same as it was under the old framework.
What Changed and What Didn’t?
On its face, the redetermination reflects the legal framework shift we identified in our earlier comment on the March 2026 Practice Notice. The Board no longer identifies an “actual invention” as a discrete analytical step. Instead, the question of what the inventor has discovered is folded into purposive construction, consistent with the Federal Court’s direction in Dusome and the Supreme Court of Canada’s framework in Free World Trust and Whirlpool. But the “actual invention” is gone in name only.
The Board’s application of the “Schlumberger question” effectively reconstitutes the very analysis the Federal Court found impermissible. After purposive construction identifies every element of the claims as essential, the Board asks what the skilled person would consider the “discovery” or “new knowledge” to be. That discovery is identified as the rules of the game. The physical cards and computer components (although the Board had found them to be essential elements of the claims) are then characterized as well-known instruments that do not supply the necessary physicality. Essential elements identified during purposive construction are discounted during the patentability assessment, not because they are inessential, but because they do not, in the Board’s view, contribute to the “discovery.”
The result is a two-step exercise that looks remarkably similar to the old one. Step one: construe the claims and identify all elements as essential. Step two: set aside the essential elements that are “well-known” and ask whether what remains (the abstracted method, algorithm, or set of rules) can independently satisfy the physicality requirement. This is the “actual invention” analysis relocated into a different stage of the inquiry.
Our Concerns, Confirmed
This outcome is consistent with what we cautioned when the March 2026 Practice Notice was released. We noted at the time that the accompanying examples suggested the Schlumberger question could function as the old “actual invention” analysis under a new label (i.e., asking what has “really been discovered” and stripping away elements that do not independently provide physicality).
The Dusome redetermination bears this out. The Board’s approach tracks the very pattern we identified: purposive construction finds all elements essential, the Schlumberger question isolates the abstract rules as the “discovery,” and the physical or computational elements are treated as unable to supply the necessary physicality. This is the same analytical move we criticized in the Practice Notice’s examples, where claims for seismic analysis, investment portfolio optimization, and neural-network-based crop irrigation were dismissed because “nothing other than an algorithm has been discovered.” The label has changed from “actual invention” to “discovery” or “new knowledge,” but the exercise (and, unsurprisingly, the outcome) remain the same.
There is a deeper conceptual difficulty with framing the inquiry around the “discovery.” Canadian patent law has long held that a discovery is not itself patentable (only an invention is). As the Supreme Court of Canada put it in Shell Oil, an invention is “the practical embodiment of the new knowledge,” and the Federal Court of Appeal in Apotex Inc v Wellcome Foundation Ltd explained that an inventor must both conceive of a new idea or discover a new thing and set that conception or discovery “into a practical shape,” such that “[m]ere conception is thus not invention unless combined with the second element of setting the idea into practical shape.” It is therefore difficult to reconcile an approach that now proceeds by asking what the “discovery” is in an effort to get at the invention. If, in assessing novelty, inventiveness, and utility, we direct the inquiry to the claimed subject matter as a whole, it is hard to see why patentability should instead be tested by stripping that subject matter down to an isolated “discovery.” This exercise, parsing the claims as the Board does, risks discarding the very elements that establish the invention.
For patentees in the software and AI spaces, the Dusome redetermination is a useful illustration that doctrinal refinement does not always translate into different results on the ground. Applicants are yet again subject to an elevated standard: identify physical elements beyond the existing essential physical elements of the invention that might have been known in the abstract or demonstrate that the invention improves the functioning of the computer itself.
Key Takeaways
1) The “actual invention” is gone in name, not in substance.
The Board has dropped the label and relocated the inquiry into purposive construction. But the analytical move (identifying the “discovery” as the abstract rules, then treating the physical or computational elements as well-known instruments that cannot supply the “something more”) is functionally the same exercise the Federal Court found impermissible. The result for the applicant is the same preliminary recommendation to refuse.
2) The Schlumberger question appears to reconstitute the old test.
As we cautioned, the Schlumberger question risks functioning as the “actual invention” analysis under a different name. The Dusome redetermination confirms this: after all elements are found essential, the Board identifies the “discovery” as the abstract rules and treats the physical or computational elements as unable to supply the “something more.” This is not a narrow comparison to the facts of Schlumberger; it is the old “actual invention” analysis, relocated and relabelled.
3) This may not be the final word.
The redetermination is preliminary. The applicant has been invited to attend a hearing (tentatively scheduled for June 16, 2026) and to provide written submissions by June 5, 2026. The applicant may also submit a further set of proposed claims. However, given the approach expressed in the Board’s preliminary views, it is likely that the Board, and by extension the Commissioner, will once again refuse this application.
4) Courts are still shaping the landscape.
We are still awaiting the Supreme Court of Canada’s guidance in Pharmascience Inc v Janssen Inc. In the meantime, the Dusome redetermination demonstrates how CIPO intends to apply its revised framework in practice. Across the Atlantic, the UK Supreme Court’s recent decision in Emotional Perception AI Ltd v Comptroller General of Patents charts a more promising course. As another leading commentator first noted here, this decision overruled the existing subject matter test for improperly importing questions of novelty into the threshold subject-matter inquiry and relocated the substantive assessment of what is new and inventive to an evidence-based, prior-art-anchored merits analysis. In our view, this is an approach more in line with the teachings of Canadian jurisprudence and one that should be instructive going forward. Given the tensions between the Practice Notice’s stated principles and the practical outcomes it produces, further judicial intervention may be necessary to give the reforms meaningful effect.