Amendments to the Patent Act, passed in 2018 to replace a prior clumsily worded provision enacted 25 years earlier, state that a patent or certificate of supplementary protection is not infringed by its being used “for the purpose of experimentation relating to the subject-matter” of the patent or certificate (Act, ss. 55.3(1) & 124(1)(f)). Regulations may indicate the factors to consider when determining what may or may not amount to experimentation, but none have been promulgated so far.

Some general comments on the new provision seem worthwhile.

• Section 55.3(1) recognizes that “[e]xcessive control by holders of … intellectual property may unduly limit the ability of the public domain to incorporate and embellish creative innovation in the long-term interests of society as a whole, or create practical obstacles to proper utilization” (Théberge v Galerie d'Art du Petit Champlain [2002] 2 SCR 336 at [32]). To allow patent holders to control downstream experimentation would be not only excessive but also counterproductive. The patent system aims to encourage innovation in the public interest without discrimination as to person: and without experimentation, innovation would be impossible.
• Section 55.3(1) should be interpreted in light of other intellectual property legislation so as to promote “harmony, coherence, and consistency” within the IP system as a whole (see CRTC Reference, 2012 SCC 68 at [37]-[38]). The section is patent law’s equivalent of the fair dealing defence for research and private study in copyright law (Copyright Act, ss. 29 to 29.2). Fair dealing is interpreted as a user’s right that requires to be liberally interpreted to promote its public interest purpose (reaffirmed very recently in Keatley Surveying Ltd v Teranet Inc, 2019 SCC 43 at [43]-[6]). Experimentation should therefore also be treated as a user right, liberally interpreted to promote patent law’s public interest purpose of encouraging innovation.
• Dictionary definitions suggest that experimentation is the action or process of experimenting: demonstrating a known fact or testing or trying something out with no certainty of outcome (e.g., Shorter OED (2000), def. “experiment” & “experimentation”). As with copyright, an act may therefore be done “for the purpose of experimentation” where it is part of a process or exercise in trying, testing or demonstrating something, as seen from the relevant user’s perspective (see the copyright approach in Pro Sieben Media AG v Carlton TV Ltd, [1999] 1 WLR 605, 614 (CA); Alta. (Education) v Cdn Copyright Licensing Agency, [2012] 2 SCR 345 at [22]-[3]).
• A right of “fair” experimentation has long existed without any special provision to that effect. Courts have achieved this by interpreting the language of the patent grant — to “make, construct, use or sell” the invention (s. 42 of the Act) – as applicable only to commercial exploitation, not fair experimental use (Micro Chemicals Ltd v Smith Kline & French [1972] SCR 506, 519–20). This principle continues but reliance on s. 55.3(1) removes any need to inquire into the use’s “fairness.” Whether the experimentation occurs for fun or profit, is fair or unfair, or is done with or without knowledge of any prior right, is irrelevant.
• The current narrow US approach to exempted research, rejected by Merck & Co v Apotex 2006 FCA 323 at [108]-[11], remains irrelevant. Canadian researchers may continue to sleep more restfully than their US counterparts (Day & Udick, “Patent Law and the Emigration of Innovation,” 94 Washington L Rev 119 (2019)).
• Section 55.3(1)’s language seems inspired by that of the UK counterpart provision based on European law, except that the Canadian section uses “for the purpose of experimentation” rather than the UK’s “for experimental purposes” (Patents Act 1977 (UK), s. 60(5)(b)). This may be a distinction without a difference or it could signal that UK jurisprudence, while of interest, is not automatically applicable to Canada. Three differences immediately stand out:
  • (a) The UK treatment of fair dealing and experimentation as derogations from the rights of IP holders, and thus exceptions to be interpreted narrowly, is inconsistent with Canada’s treatment of similar limitations (Comm’r of Patents v Farbwerke Hoechst [1964] SCR 49, 57) and the liberal treatment of fair dealing as a user right, earlier noted.
  • (b) Canada’s former statutory experimental defence, now repealed, allowed working “solely” for an experimental purpose. The elimination of the word “solely” by s. 55.3(1) is significant. Experimentation need now be only “a” purpose of the activity (the French version similarly uses “un but”). Mixed purposes are therefore acceptable, contrary to UK authority which adds qualifications such as “preponderant” or “reasonably necessary” to the experimental purpose (compare Corevalve Inc v Edwards Lifesciences AG [2009] EWHC 6 at [79]-[81]; Meter-Tech LLC v British Gas Trading Ltd [2016] EWHC 2278 at [231]-[6]).
  • (c) UK courts have interpreted “relating to” in the phrase “experimental purposes relating to the subject-matter of the invention” as if it meant the same as “on” the former invention: “a real and direct connection” between the purpose and the invention is required (Auchincloss v Agricultural & Veterinary Supplies Ltd [1999] RPC 397 at 405-6). In Canada, the phrase “relating to” is among “the widest of any expression intended to convey some connection between two related subject matters” (Slattery (Trustee of) v Slattery,[1993] 3 SCR 430 at 445, emphasis added). Reading words into s. 55.3(1) so as to require a “substantial” or “direct”, instead of merely “some”, connection is therefore unwarranted.
• The issue of experimentation also arises in another context: when do experiments, whether by an inventor or third party, disclose an invention by making it “available to the public” (Act, s. 28.2(1)(b)) and thus unpatentable? It has for pragmatic reasons been held that experiments to perfect an invention or test its merits or practical value do not make it available to the public until they cease being “reasonable and necessary” (Canadian Patent Scaffolding Co v Delzotto Enterprises Ltd (1978), 42 CPR (2d) 7 at 24 (1980), 47 CPR (2d) 77). Such experiments should equally qualify as experimentation under s. 55.3(1) even where no longer “reasonable and necessary”: for, again, no such qualification appears in s. 55.3(1) and none should be implied. For s. 55.3(1)’s purposes, experimentation should not be minutely examined to see whether its exercise, in hindsight, seems too little or too much. To be of practical value, the experimentation right should protect bad and good experimenters alike. Section 55.3(1)’s shield lifts only when the invention enters commerce by moving from the laboratory bench to the factory or showroom.
• Where two or more parties are involved — e.g., supplier and buyer — their separate positions should be considered: one, both, or neither may be entitled to the defence. For example, a seller may supply samples commercially for a buyer to test for suitability for or adaptability to his purposes. The facts may reveal both parties’ or only the buyer’s overall purpose to be experimentation. If the buyer decided that a use or adaptation was suitable, any further use would then fall outside s. 55.3(1).

Written by David Vaver, Professor of Intellectual Property Law, Osgoode Hall Law School