Conflict of jurisdiction and the internet – a split Supreme Court of Canada leaves us hanging

In the highly anticipated Haaretz.com v Goldhar decision, the Supreme Court of Canada split five ways in its approach to grappling with multijurisdictional claims of internet-based defamation. 

Goldhar, a Canadian businessman, sued Israeli newspaper Haaretz for writing an article criticizing Goldhar’s management style in his Canadian business and his ownership and management of an Israeli soccer team. Around 70,000 people had read the online Haaretz article in Israel, while around 200-300 people had read it in Canada. 

Harretz filed a motion to stay the action for lack of jurisdiction or alternatively that Israel was the more convenient forum. The motion judge dismissed the motion finding that Ontario courts had properly taken jurisdiction and because Goldhar was not seeking damages for reputational harm suffered in Israel, Ontario was clearly the more appropriate forum. The Ontario Court of Appeal concluded that while the motion judge had made some legal errors, the errors did not significantly affect the decision.

In a five-way split decision, the Supreme Court of Canada found that Israel was the more appropriate forum for the defamation claim, even though Ontario had jurisdiction simpliciter.

What is (jurisdiction) simpliciter is not always conveniens

While there was consensus that Ontario had jurisdiction simpliciter, the court splintered on how to address the forum non conveniens analysis in addressing online defamation claims. The key division related to whether the ease with which jurisdiction simpliciter could be established in such claims required a more robust forum non conveniens analysis to balance potential unfairness.

The governing decision by Côté J. (Brown and Rowe JJ. concurring) confirmed that the real and substantial connection test governs jurisdiction simpliciter, even in the context of internet defamation where finding a connection will be virtually automatic.

To balance the automaticity of establishing a connection, Côté J. found that the forum non conveniens analysis must be approached at a “robust and carefully scrutinized review” in internet defamation cases¹. In Côté J.’s view this standard may alleviate some of the potential unfairness that arises with the ease of meeting jurisdiction simpliciter. In separate reasons, Karakatsanis, Wagner, and Abella JJ all agreed, cementing this standard as governing law.

While the majority’s reasons state that adopting this new language does not impose a new threshold for considering forum non conveniens in internet defamation cases, the dissent written by McLachlin CJC (Moldaver and Gascon JJ concurring) took the view that using such novel language—the “robust and carefully scrutinized review” rather than the traditional language of “clearly more appropriate”—did reflect a novel approach.

In the view of the dissent, such a change was unnecessary, despite the reality that presumed jurisdiction will be virtually automatic. It noted the presumption can still be rebutted if there is only a weak relationship between the action and the forum or where there was not reasonable foreseeability that a claim would proceed in that jurisdiction. Properly applied, this could balance any potential unfairness, and applying an elevated standard was not necessary, and would undermine stability and increase costs and uncertainty for parties.

Defamation claims and “the most substantial harm”

In addressing the forum non conveniens analysis, the court was most divided on the thorny issue of what law governed the defamation claims and what role the governing law played within the larger analysis.

Côté J. confirmed the use of the lex loci delicti (the place where the tort occurs) rule, while cautioning that little weight should be placed on the applicable law factor when establishing forum non conveniens, given that each forum would normally apply its own law to the matter, and therefore the applicable law would not affect whether it would be more fair or efficient in an alternate forum. Both the dissent and Karakatsanis J. affirmed the use of the lex loci delicti rule, but differed from Côté J. on the role of governing law within the broader forum non conveniens test.

Abella and Wagner J.J. differed, proposing that governing law be based on where “most substantial harm to the plaintiff’s reputation occurred,” drawing from Australian and European jurisprudence (Abella J. went one step further, leaving an open invitation for future courts to adopt the test to govern jurisdiction simpliciter in such claims).

Underpinning this proposal was the aim of narrowing, in a rational way, the potential geographic breadth of internet defamation claims. Since the essence of harm in defamation is damage to the reputation, Abella J. reasoned that the framework for determining forum non conveniens (and jurisdiction) should focus on where the plaintiff suffered the most substantial harm to his or her reputation. However, this novel approach did not find favour with the remaining members of the court.

While the opinions varied widely, the net result is modest: lex loci delicti will remain the test for governing law in defamation claims, and should be approached as before by considering only the law in the home court, but should only be given limited weight within the forum non conveniens analysis.

Slow change in a fast internet world

Both jurisprudential disputes arose from the recognition that online torts frequently occur concurrently in multiple jurisdictions. By relying on the place of the tort both to establish jurisdiction and establish governing law, the pre-existing law has opened the door to forum shopping in defamation claims—but also to other claims where the elements of a tort occur simultaneously in many geographic jurisdictions. While there was little agreement between the members of the court, there appears to be a clear recognition that something about jurisdiction and the internet requires a change in our law.

It will now be the lower courts’ responsibility to decide if the “robust and carefully scrutinized review” adopted by Côté J. is simply new language for the same forum non conveniens analysis, or if adopting this new language creates a new threshold that will reduce the ability of litigants to engage in libel tourism, as suggested by the dissent.

It seems likely that new language will lead to a new test—one more likely to halt claims with only tenuous connections to Canada. Unfortunately, this issue remains far from certain.

The authors would like to thank Brenna Epp, articling student, for her assistance preparing this legal update.

Footnote

¹ At para. 48.