“Inducing” Copyright Infringement in Canada: Is it a Thing?

“Inducing” Copyright Infringement in Canada: Is it a Thing?


David Vaver headshotProf. David Vaver is a member of IP Osgoode's Advisory Board, an Intellectual Property Law Professor at Osgoode Hall Law School and an Emeritus Professor of IP & IT Law at the University of Oxford.


A version of this paper is scheduled to be published in the Intellectual Property Journal. Thanks to Ronald M. Lieberman, Christopher G. Moore, and Maxine Vaver for helpful comments on an earlier draft. They of course share no responsibility for what appears here.

Is there such a thing as “inducing” copyright infringement? There isn’t in the United Kingdom, but the Federal Court in Bell Canada v L3D Distributing Inc (INL3D) 2021 FC 832 (“L3D Distributing”) thought there was in Canada and the defendants had done it. Indeed, the court thought that inducing infringement, which to date had been considered wrong only in respect of patents, applied to all forms of intellectual property (“IP”).

In Compo Co Ltd v Blue Crest Music Inc [1980] 1 SCR 357, 372, the Supreme Court said that “[c]ourts in this technical field of copyright have found it prudent to make their judicial answers congruent with the legal issues raised in the pro­ceeding at hand leaving, so far as possible, analo­gies, examples and hypothetical questions to another day.” This caution applies particularly where the case is undefended, as L3D Distributing was, and the plaintiffs are asking the Court to make new law. Where a case would succeed on more familiar grounds, it is prudent simply to record the additional plea or treat it as raising a hypothetical question; for what may look rock solid when presented by one side often crumbles once the other side is put. As Justice Megarry said when repudiating a view he had expressed in his standard land law treatise, it is dangerous to reach a settled conclusion “without the aid of the purifying ordeal of skilled argument on the specific facts of a contested case” (Cordell v Second Cranfield Properties Ltd [1969] Ch 9, 16, emphasis added). Audi alteram partem is, after all, prudence epitomised.

It thus seems best to treat the Federal Court’s views on inducement as an indication that the point was arguable, rather than necessarily sound, especially since such a holding was not strictly necessary. The defendants had been retailing pre-loaded set-top boxes that gave viewers unauthorized access to television broadcast and cable programming in which the plaintiffs owned or controlled the copyright. The Court unexceptionably found the defendants liable on three counts other than inducement: (a) making copyright-protected work available for telecommunication to the public; (b) authorizing such communication by telecommunication (both contrary to the Copyright Act RSC 1985, c C-42); and (c) selling equipment that decrypted cable transmissions (contrary to the Radiocommunication Act RSC 1985, c R-2) — none of which the plaintiffs authorized. Neither the inducement count nor the radio-communication count added anything material to counts (a) and (b): the Court thought the latter two counts themselves enough to sustain the grant of an injunction and a $30 million award of statutory and punitive damages and costs against the defendants.

So is the Court’s view, that inducing copyright or IP infringement is indeed a Thing, sound? I suggest not.

One must of course first start with the Copyright Act itself and deal with what the Supreme Court also said in Compo (above):

“Mr Hughes … put it very well when he said that copyright law is neither tort law nor property law in classifi­cation, but is statutory law. Copyright legislation simply creates rights and obligations upon the terms and in the circumstances set out in the statute. … It does not assist the interpretive analysis to import tort concepts. The legislation speaks for itself…” ([1980] 1 SCR at 372-3).

So, if we take the Supreme Court literally, since the Act defines copyright infringement as doing “anything that by this Act only the copyright owner has the right to do” (s. 27(1)), and “inducement” is — unlike making available and authorization — not something only the copyright owner has the right to do, the inquiry should stop right there: inducement cannot be a Thing because the Act doesn’t say it is.

The trouble is that the Court’s statement in Compo, despite being parroted in countless copyright cases, cannot be taken literally; for in its very next paragraph, the Court approves decisions of the UK and US courts that hold authorization generally, and authorizing the making of sound recordings specifically, to be “separate torts” (the Court’s emphasis), if done without the copyright owner’s consent. So, according to the Court, the interpretation of the Copyright Act is not assisted by importing tort concepts into something the Court chooses to call a “tort.” The authority for this method of reasoning is presumably Justice Humpty Dumpty’s well-known dicta in Lewis Carroll’s Through the Looking Glass.

In fact, it is well settled that tort concepts of accessory liability — the civil law equivalent of aiding and abetting in criminal law — do apply to the Copyright Act, and indeed all IP Acts, even where the legislation is silent on the point. How else are corporations found vicariously liable for IP infringements committed by their employees? How else are directors held liable for any egregious conduct in connection with their company’s IP infringements? All legislation that creates a statutory tort or wrong is presumed to make accessories liable the way they are in tort, unless the legislation expressly or impliedly excludes such liability (Majrowski v Guy's & St Thomas' NHS Trust [2006] UKHL 34 at [16], [72], [74]). IP infringements are one such example of a statutory tort, as Compo and many other courts in Canada and elsewhere have long recognized. Accessories are considered themselves to infringe IP jointly and severally by participating in another’s infringement. They are thus subject to the full panoply of infringement remedies, subject to whatever is recovered from the primary infringer. The most common accessory is the joint tortfeasor or joint infringer, who is indeed specifically contemplated by the Copyright Act (see, e.g., sections 17(2) (right of remuneration), 38.1(1) (statutory damages), 44.07(4) (charges for border stoppage)).

Some academic writing argues for a wider tort principle of accessory liability: anyone who contributes knowingly and substantially to another’s infringement by helping, encouraging, or inducing them to infringe should be liable to the IP holder for the full panoply of monetary, injunctive, and other equitable remedies: see, e.g., Paul Davies, Accessory Liability (Oxford: Hart 2015), passim. US courts have gone this far (MGM Studios Inc v Grokster Ltd 545 US 913 (2005)) but the UK Supreme Court has not. Concerned to keep accessory liability within reasonable bounds, the Court has insisted on the additional requirement that the wrongdoing must be found to be joint before monetary remedies can be imposed on defendants. The combination of “active co-operation and commonality of intention” then helps avoid “constrict[ing others’] freedom to engage in activities which are otherwise lawful” (Sea Shepherd UK v Fish & Fish Ltd [2015] UKSC 10 at [44], also [56] & [58]; similarly Google LLC v Defteros [2022] HCA 27 at [204], [212] (Aust HC) (defamation)). In other words, inducing copyright or other IP infringement is a Thing only if, additionally, a common design or intent between an inducer and inducee to do acts that amount to infringement is found to exist, so that they become joint infringers (CBS Songs Ltd v Amstrad Plc [1988] RPC 567, 606-9 (HL)).

Joint infringement is a flexible enough concept to encompass even the financing of infringing activities, as occurred in the Sea Shepherd case (above). But flexibility can create uncertainty and division and much will depend on the view a judge takes of the facts of a case (Rutman v Rabinowitz 2018 ONCA 80 at [31]-[5]). Thus in Sea Shepherd, while agreeing on the legal principles of joint liability, the UK Supreme Court divided on their application. All accepted the presence of a common design between Sea Shepherd UK and its overseas parent organisation, Sea Shepherd Conservation Society (“SSCS”), under which the UK company supported SSCS in its attempts to stop bluefin tuna fishing in the Mediterranean. A majority of the Court, however, held that the UK company’s local fundraising efforts for SSCS’s activities were too insignificant to have induced the trespass that SSCS committed in attacking the boats that were hauling tuna to the plaintiff’s fish farm off Malta.

Joint liability is largely about finding someone worth suing for another entity’s sometimes more substantial wrongdoing — someone with the wherewithal to pay compensation or hand over any wrongfully made gains. That was what the fish farm or its insurers in Sea Shepherd were ultimately after. Sea Shepherd UK’s assets might not be enough to cover the plaintiff’s claimed losses and anyway it had not carried out the attack. The UK company could, however, be used as a way to get the English courts to take jurisdiction over the better-funded foreign-based SSCS and then hold both companies jointly and severally liable. The tactic would have succeeded had the UK company influenced SSCS’s activities more. The UK company would then have been a participant in a tort partly committed in England, even though the major wrongdoing occurred abroad. As it turned out, however, the finding of no joint tortfeasance meant that the plaintiff could not proceed against either defendant and had to find some other way or place to recoup its losses.

Behaviour that is not serious enough to rise to the level of joint infringement can nevertheless have consequences. The person may not have to pay money but they may have to adjust their conduct. Courts everywhere have proved willing to deal with intermediaries such as online marketplaces and search engines that facilitate or enable IP infringement while not technically being infringers or accessories. Thus, a British Columbian court issued an injunction against Google ordering it to de-index websites worldwide that offered specified IP-infringing goods, and was upheld on appeal (Google Inc v Equustek Solutions Inc [2017] 1 SCR 824). Google was no infringer or accessory but it at least owed the IP holder a duty to take reasonable steps to prevent such infringements on its platform as were brought to its attention. It was liable to pay money only if the injunction was ignored.  

Canadian case law has not adopted a generalized tort principle of accessory liability beyond joint infringement. The IP cases reflect this by refusing to label as infringer anyone who merely helps, facilitates, or contributes to another’s infringement (e.g., Copeland-Chatterson Co v Hatton (1906) 10 Ex CR 224, 241, aff’d without appealing this point (1906) 37 SCR 651 (patent); Nycomed Canada Inc v Teva Canada Ltd 2012 FCA 195 at [3], aff’ing 2011 FC 1441 (patent); Prest-O-Lite Co v People's Gas Supply Co (1917) 16 Ex CR 386, 389, aff’d (1917) 55 SCR 540 (trademark)). Such a person may, however, owe IP holders a duty to take steps, where feasible, to prevent infringements IP holders bring to its notice, although without any monetary liability to them except where any injunction is disobeyed.

Whether Canadian IP policy should change tack and hold that inducement or contribution, without more, creates monetary liability — as well as the remaining panoply of remedies — for infringement is thus a decision more suited to legislative amendment than to a first instance court in an undefended case. Parliamentary drafters know how to extend such liability where that is intended, and IP rights are, after all, “derived from statute and not from the Ten Commandments” (CBS Songs, above, at 608).

There is, admittedly, one exception to this position, and it is what the Federal Court in L3D Distributing relied on. The Court’s predecessor, the Exchequer Court of Canada, held over a century ago that knowingly inducing  infringement of a patent, without more, itself infringes the patent and subjects the inducer to the full range of infringement remedies, where the inducement works (Copeland-Chatterson, above, at 92). A line of Federal Court and Court of Appeal decisions since then has consolidated this position for patents (see, e.g., Corlac Inc v Weatherford Canada Inc 2011 FCA 228 at [162]ff; Western Oilfield Equipment Rentals Ltd v M-I LLC 2021 FCA 24 at [48]-[70]), eclipsing (until recently) pleas of joint infringement by common design (see now Rovi Guides Inc v Videotron Ltd 2022 FC 981 at [46]-[7]). These decisions on inducement are probably now beyond question, even though the Supreme Court has never examined the issue. Yet a major weakness runs through them: the key English patent case on which Copeland-Chatterson (above) relied, Innes v Short (1898) 15 RPC 449 (QB), is not treated in England as holding that inducement is itself wrongful. Rather, in line with the general law of torts, Innes v Short has been authoritatively interpreted there as holding that the inducement must be in furtherance of a common design, so that the inducer and the inducee are joint infringers (see CBS Songs,  above at 606-7; similarly Sea Shepherd, above at [23]). Without the presence of a common design, Innes is otherwise a case of “somewhat doubtful authority” (Adhesive Dry Mounting Co Ltd v Trapp & Co (1910) 27 RPC 341, 353 (Ch); see too Re L’Air Liquide SA’s Patent App’n (1930) 49 RPC 428, 430; Firth Industries Ltd v Polyglas Engineering Pty Ltd [1975] HCA 25 at [16] (Aust HC); CCOM Pty Ltd v Jiejing Pty Ltd (No 2) [1993] FCA 653 at [153], aff’d [1994] FCA 1168 (Aust Full Fed Ct)), even though the judge there decided to award only an injunction but no monetary remedy against the infringer.

The Federal Court in L3D Distributing went into none of this. It simply applied the syllogism that what held for Canadian patent law held for copyright and indeed all Canadian IP. Inducing infringement was therefore, without more, wrongful (L3D Distributing at [71]-[4]). This conclusion, however, rests on shaky ground. Since inducement liability is not found in the Copyright Act, the first question must be whether the Act includes it by implication. That in turn depends on the answers to two further questions: Is inducement a general principle of accessory liability in tort? If so, is its inclusion consistent with the Act?

The most likely answer to both these questions is, no. Since inducement is not a general tort law principle in Canadian law, it is a large leap to add it as such to copyright law, or indeed any non-patent IP. The patent law anomaly need not infect other IP. And even if inducement were such a general principle, its inclusion seems inconsistent with the Copyright Act. Section 27 sets out in some detail who infringes what copyright and when. No room seems left for another form of infringement, even more so because someone who induces infringement necessarily “sanctions, approves, and countenances” it and thus falls within the Act’s authorization right (SOCAN v Entertainment Software Ass’n 2022 SCC 30 at [104]). Authorization indeed elevates inducement to primary rather than accessory liability: authorization is complete whether or not any later infringement occurs (SOCAN, ibid at [104]-[5], [107]). The greater liability (authorization) must thus encompass the lesser (inducement).

The undefended nature of the case before the Federal Court highlights further concerns with its conclusions. First, on copyright: (i) inducement liability is found in no other Canadian copyright decision in the century since the Copyright Act was enacted; and (ii) the Court took no account of the unanimous contrary holding of the UK’s highest court on comparable copyright provisions, that inducement is only a form of joint infringement (CBS Songs, above, at 608-9, foll’d as a general tort principle in Southpaw Credit Opportunity Master Fund LP v Asian Coast Development (Canada) Ltd 2013 BCSC 187 at [200], [205]).

Second, on non-patent IP generally: (i) the Court’s comments here were clearly obiter; (ii) the Court did not consider the contrary views proffered in the leading UK decision, that inducement as a form of joint infringement applies to all IP (Sea Shepherd, above at [44], [56], [58]); (iii) the only non-copyright authority the Court cited (Hanson Int’l Inc v Whirley Industries Inc 2002 FCT 1045) was an industrial design case where a prothonotary’s refusal to strike an assertion of inducement before trial as being unarguable is clearly no authority that such an assertion would succeed; (iv) the Court’s statement that inducement is a “common law cause” that is “[i]n addition to the statutory causes of action” (L3D Distributing, above at [71]) is an odd comment for a court that lacks jurisdiction over any “common law cause;” (v) inducement is anyway not a “common law cause” (McCain Foods Ltd v J R Simplot Co 2021 FCA 4 at [62] (“not a distinct tort”)), but at best only a form of accessory liability.

Had joint copyright infringement been properly pleaded in L3D Distributing, the defendants would likely have been found to be joint infringers with their consumers. The necessary common design or intent between inducer and inducee need not be express. A tacit agreement, inferable from the circumstances of even an undefended case, suffices. Thus UK courts have found joint infringement based on a common design or intent where (i) the inducer was a peer-to-peer file-sharing operation or an app using BitTorrent or some other means of circumventing encrypted or password-protected content, and (ii) the targets were users or hosting websites that were induced or persuaded to stream, download or make infringing content available in furtherance of the common design (Dramatico Entertainment Ltd v British Sky Broadcasting Ltd [2012] EWHC 268 at [82]-[3] (Ch) (Pirate Bay); Twentieth Century Fox Film Corp v Sky UK Ltd [2015] EWHC 1082 at [51]-[6] (Ch) (app using BitTorrent protocol); Young Turks Recordings Ltd v British Telecommunications Plc [2021] EWHC 410 at [68]-[71] (Ch) (“stream ripping” app). Since a finding of joint infringement is both “fact sensitive and policy-driven” (Barclay-Watt v Alpha Panareti Public Ltd [2022] EWCA Civ 1169 at [62]), it is no great leap in L3D Distributing to infer not only inducement but also a common design to infringe copyright between the sellers and buyers of the equipment. Everyone intended that it be used to access programming which all knew was lawfully available only through other services authorized by the copyright owners of the programming. All roads can lead to Rome, provided one avoids the ditches.