Table of Contents

A prior art search Montreal litigation counsel can rely on has to match where the fights actually start — and in Quebec’s largest city that means aerospace, artificial intelligence, pharmaceuticals and video games. Montreal conducts roughly three-quarters of Canada’s aerospace research and development, anchored by Bombardier, Pratt & Whitney Canada and CAE, and it has become one of the world’s densest AI-research clusters around Mila, the Quebec Artificial Intelligence Institute. When those portfolios are asserted, patent validity is decided by the Federal Court of Canada, which alone can strike a patent down with effect against the world. PerspireIP builds invalidity-grade searches for the parties challenging those patents before the Canadian courts.
Where a Montreal patent case is actually heard
Patent law in Canada is a matter of federal competence, and that shapes every Montreal dispute. The Federal Court of Canada has jurisdiction across the whole country over infringement and, critically, over impeachment — an action to have a patent or any of its claims declared invalid or void. Under the Patent Act, an impeachment action can be brought by the Attorney General of Canada or by any interested person, and a Montreal-based defendant or challenger files in the Federal Court’s Montreal registry rather than in a local Quebec court.
The choice of forum decides how far an invalidity win reaches. Only the Federal Court can declare a patent invalid in rem — void and unenforceable against all third parties, with the register corrected accordingly. The Quebec Superior Court, as a provincial superior court, can hear a patent infringement claim, but it cannot strike a patent from the register; any finding of invalidity there binds only the parties to that case (in personam).
In practice almost all Canadian patent litigation runs in the Federal Court, precisely because a defendant wants an invalidity ruling that ends the patent for good rather than one that protects only itself. From there, appeals go to the Federal Court of Appeal, and then, with leave, to the Supreme Court of Canada. Prior art scoped to a Federal Court impeachment action — charted, dated and trial-ready — is what carries that fight.
Federal Court procedure also shapes how prior art is deployed. Unlike the bifurcated German model, validity and infringement are tried together in a single proceeding, so an invalidity attack must be complete from the pleadings onward. Cases run through documentary and oral discovery, exchange of expert reports, and a trial where those experts are cross-examined on the art. References that are charted to the claim elements and firmly dated survive that scrutiny; loosely sourced art rarely does. The court’s Montreal registry and its case-management judges keep these matters on a tight schedule.
- Federal Court of Canada — national jurisdiction over infringement and impeachment; the only court that can invalidate a patent in rem
- Quebec Superior Court — a provincial superior court that can hear infringement but cannot strike a patent from the register; invalidity there binds only the parties
- Federal Court of Appeal, then the Supreme Court of Canada — the appellate route on validity
- In rem vs in personam — why most challengers choose the Federal Court to kill a patent for everyone, not just themselves
The invalidity grounds a prior art search Montreal must target
Under the Patent Act a claim can be attacked on several grounds, and the ones that turn on prior art are where an invalidity search earns its keep. The two heavyweights are anticipation (lack of novelty) and obviousness (lack of inventive step), both governed by the Supreme Court of Canada’s decision in Apotex v Sanofi-Synthelabo, 2008 SCC 61 — the Plavix case. A prior art search Montreal defendants can build a case on has to be scoped to those tests from the outset.
Anticipation in Canada is a two-part test. A single earlier reference must satisfy prior disclosure — directing the skilled person, in every case and without possibility of error, to the claimed invention — and enablement, meaning that person could perform the invention without undue burden. A reference that discloses but does not enable, or enables only through further inventive steps, does not anticipate. That is a demanding standard, and it puts a premium on finding the one document that does both.
Obviousness follows the four-step Sanofi framework: identify the skilled person and the common general knowledge, define the inventive concept of the claim, identify the differences between the prior art and that concept, and ask whether those differences would have been obvious. Where the field invites experimentation, the court applies an “obvious to try” analysis, weighing whether success was self-evident, the effort required, and the motivation in the art. Both branches depend on a full, well-dated map of what was already public.
Two further grounds do not turn on a search but often ride alongside one. Overbreadth asks whether the claims are wider than the invention actually made or disclosed; the Federal Court of Appeal has confirmed it is a distinct ground that must be assessed on its own. Insufficiency asks whether the specification teaches a skilled person how to work the invention across the full scope claimed. A strong prior-art record frequently sharpens both, by exposing how far a claim reaches beyond what the disclosure genuinely enables.
- Anticipation — one earlier disclosure that both discloses and enables the claim (Sanofi two-part test)
- Obviousness — the claim is obvious over the prior art, often a combination, under the four-step Sanofi framework and “obvious to try”
- Overbreadth — the claims are broader than the invention made or disclosed; the Federal Court of Appeal treats this as a distinct ground
- Insufficiency — the specification does not describe the invention well enough for a skilled person to work it
Utility after AstraZeneca: the promise doctrine is gone
For more than a decade, Canadian patents — pharmaceutical patents especially — were vulnerable to the promise doctrine: if the disclosure was read as promising a particular result, the whole patent could fall for lack of utility unless that promise was demonstrated or soundly predicted at the filing date. Dozens of patents were struck down on this basis, and it made a careful reading of the specification’s stated uses a live invalidity strategy.
That changed in 2017. In AstraZeneca Canada Inc v Apotex Inc, 2017 SCC 36 (the Nexium/esomeprazole case), the Supreme Court of Canada held that the promise doctrine “is not good law” and is not the correct way to assess utility under section 2 of the Patent Act. A scintilla of utility — a single use related to the subject-matter of the claim — now satisfies the requirement.
The practical effect for a Montreal challenger is a shift of weight. With utility no longer the soft target it once was, invalidity campaigns lean harder on anticipation, obviousness, overbreadth and insufficiency — the grounds that live or die on prior art. That is exactly why a rigorous, date-proven search matters more in Canada today than it did before 2017: the winning attack is almost always evidentiary, not a matter of parsing the patentee’s own promises.
Aerospace prior art: Montreal’s engineering docket
Montreal is one of the world’s three great aerospace capitals, and it generates a deep, highly technical body of patent disputes. Bombardier designs and builds business and regional aircraft here; Pratt & Whitney Canada develops turboprop, turbofan and turboshaft engines from Longueuil; CAE is a global leader in flight simulation and training; and Bell Textron Canada anchors the region’s rotorcraft base. Roughly three-quarters of Canada’s aerospace R&D happens in and around the city.
The patents that get asserted in this world are dense: turbine and combustion architectures, gearbox and transmission designs, additive-manufacturing methods, advanced alloys and composites, avionics, flight-control software and simulation systems. When a supplier or a rival is sued over one of these, the decisive prior art frequently sits outside the patent literature — in engineering standards, SAE and ISO papers, technical reports, airworthiness documentation and older product literature that examiners never saw.
Invalidity work in aerospace therefore leans heavily on non-patent retrieval and on proof of public availability. The strongest reference is often a conference paper, a defence or space-agency technical report, or a supplier bulletin whose exact publication date decides whether it counts as prior art at all. A prior art search Montreal aerospace defendants can rely on mines that grey literature and dates every reference to the day, so the court sees a disclosure the patentee’s prosecution never accounted for.
AI and software prior art: the Mila cluster
Montreal is a global centre of gravity for artificial intelligence. Mila, the Quebec Artificial Intelligence Institute, sits at the heart of a dense academic and industrial ecosystem — the legacy of Element AI, plus major research labs run by global technology companies and a thick layer of startups. That concentration produces a fast-moving stream of machine-learning, computer-vision and natural-language patents, and, when they are asserted, a stream of accused implementers who need to show the claimed method was already known.
AI and software prior art has its own evidentiary character. The most damaging references are rarely patents; they are arXiv preprints, NeurIPS, ICML and ICLR conference papers, university theses, open-source code repositories and technical blog posts. In a field that moves in months, the priority date of the asserted claim is unforgiving, and the whole fight often turns on proving that a preprint or a code commit was publicly available before that date.
That is where searching shades into forensics. We chase version histories, repository commit logs, mailing-list archives, conference proceedings and cached pages, and we document the public-availability date of every reference so it cannot be waved away. Because so much AI art is enabling by nature — a paper that both describes and teaches the technique — it feeds anticipation and obviousness attacks alike under the Sanofi tests.
Pharma, gaming and a Quebec branding note
Two more Montreal industries round out the docket. The city has a long life-sciences and pharmaceutical presence, and pharma is the field where the collapse of the promise doctrine matters most: invalidity now hinges on anticipation and obviousness over the medicinal-chemistry and clinical literature, so genus/species disclosures, prior compounds and dosage-form art have to be searched with real rigour. Selection-patent and “obvious to try” arguments, both shaped by Sanofi, are common here.
Montreal is also one of the planet’s largest video-game development hubs, anchored by Ubisoft Montreal, Behaviour Interactive, EA and WB Games. Game-technology patents — rendering, matchmaking, animation, netcode and interface methods — are increasingly asserted, and the prior art lives in game engines, GDC talks, developer documentation, patches and older shipped titles. Establishing what a released build actually disclosed, and when, is its own searching discipline.
One local practicality is worth flagging, even though it is not a validity issue. Quebec’s Charter of the French Language, tightened by Bill 96, imposes French-language requirements on commercial signage, packaging and trademarks — with the final provisions in force since 1 June 2025. That affects branding and marketing rather than whether a patent is valid, but it is part of the operating landscape any company litigating in Montreal should keep on its radar.
How PerspireIP builds a Montreal invalidity search
Every engagement follows the same disciplined path. We map the asserted claims element by element, fix the priority date that actually governs each one, and search against that date rather than the filing date printed on the cover. For aerospace, AI, pharma and gaming assertions we run patent searching alongside deep non-patent retrieval — engineering standards, technical reports, conference proceedings, preprints, code repositories, theses and archived product documentation — and we date every reference to the day so its status as prior art cannot be disputed.
- Claim charting mapped to the Canadian tests — anticipation (disclosure plus enablement) and obviousness (the four-step Sanofi framework and “obvious to try”)
- Deep non-patent retrieval across engineering, standards, AI, life-sciences and games sources, in English and French
- Public-availability dating for every reference, evidenced for journals, conference papers, preprints and code commits alike
- Prior art scoped to a Federal Court of Canada impeachment action or an invalidity defence, and to overbreadth and insufficiency as well as prior-art grounds
- A written invalidity analysis and reference packages ready for Federal Court counsel and their experts
We work alongside your Canadian and international counsel as a specialist search partner, deliver to Federal Court deadlines, and keep every engagement confidential. Whether you are an aerospace supplier facing an OEM’s assertion, a technology company challenging an AI or software patent, a generic pharmaceutical maker, or litigation counsel preparing an invalidity defence, we scale to fit — a single search, a multi-patent campaign or ongoing portfolio support. Send us the patent number and your key dates, and we will scope a prior art search Montreal project within one business day.
IP Landscape & Resources in Montreal
Key intellectual-property authorities and venues relevant to Montreal:
- CIPO (Canadian Intellectual Property Office) — the office that grants Canadian patents, whose validity is decided by the Federal Court
- Federal Court of Canada — the court with national jurisdiction over patent infringement and impeachment, and the only court that can invalidate a patent in rem
- Patent Act (Justice Laws Canada) — RSC 1985 c. P-4, the federal statute setting out impeachment actions and the grounds on which a Canadian patent can be declared invalid
- WIPO Lex โ Canada — WIPO's consolidated collection of Canada's intellectual property laws, including the Patent Act and its regulations
Request a Prior Art Search in Montreal
Request a Prior Art Search in Montreal
Get an invalidity-grade prior-art search built for a Federal Court of Canada impeachment action or invalidity defence, tuned for aerospace, AI, pharma and gaming claims and scoped to the Sanofi anticipation and obviousness tests. Send us the patent number and your key dates, and we will scope the work within one business day.
Explore related PerspireIP services: Prior Art Litigation Search · Patent Invalidation · Patent Infringement Analysis.
Frequently Asked Questions
Which court hears a Montreal patent case?
Patent litigation in Montreal is almost always heard by the Federal Court of Canada, which has national jurisdiction over both infringement and impeachment (invalidity) actions under the Patent Act. The Federal Court is the only court that can declare a patent invalid in rem — void and unenforceable against everyone, with the register corrected. The Quebec Superior Court, as a provincial superior court, can hear an infringement claim, but it cannot strike a patent from the register, and any invalidity finding there binds only the parties to that case. Appeals go to the Federal Court of Appeal and then, with leave, to the Supreme Court of Canada.
How do I invalidate a patent asserted in Montreal?
You bring an impeachment action in the Federal Court of Canada, or raise invalidity as a defence and counterclaim inside an infringement suit. The prior-art grounds are anticipation (lack of novelty) and obviousness (lack of inventive step), both governed by the Supreme Court’s Sanofi decision, alongside overbreadth and insufficiency. Anticipation requires a single reference that both discloses and enables the claim; obviousness follows the four-step Sanofi framework, including an ‘obvious to try’ analysis where the field invites experimentation. A claim-charted invalidity search, dated to each claim’s priority date, is what carries the case.
Does the old ‘promise doctrine’ still threaten Canadian patents?
No. In AstraZeneca Canada Inc v Apotex Inc, 2017 SCC 36, the Supreme Court of Canada held that the promise doctrine is not good law and is not the correct way to assess utility under section 2 of the Patent Act. A single use related to the subject-matter of the claim now satisfies the utility requirement. Because utility is no longer the soft target it once was, invalidity campaigns in Montreal now lean harder on anticipation, obviousness, overbreadth and insufficiency — the grounds that depend on a rigorous, well-dated prior-art search.
What kinds of patents get litigated in Montreal?
Montreal’s docket is driven by its industries. The aerospace cluster around Bombardier, Pratt & Whitney Canada, CAE and Bell generates disputes over engines, avionics, materials and simulation, where the decisive art often sits in engineering standards and technical reports. The AI cluster around Mila produces machine-learning disputes where the key prior art is arXiv preprints, conference papers and code repositories. Pharmaceutical and video-game patents round out the mix. For accused parties in each field, a specialist invalidity search focused on non-patent literature and precise public-availability dating is usually what makes the difference.