Image Alt

The Singapore Law Gazette

Choice of Law for Torts in the Virtual World and the Future of Virtual Dispute Resolution

What law applies to torts and non-contractual wrongs in the online virtual world when the parties and acts traverse different jurisdictions? How does the choice of law rules operate? What novel approaches to this area of law and to online dispute resolution are possible in the horizon? This article explores these questions.


Let us imagine that someone in the metaverse or on the Internet did a wrong to you (outside of a contractual relationship), e.g. defamed you, cheated you, or passed off as your business. That someone is from another country A. The act had effects in yet another country B. Or the act was done such that technically, the relevant communications were published in a physical server in yet another country C. What law would apply to your claim?

While there may be choice of law clauses in the terms of use of the metaverse service or the virtual platform, these apply to the relationship between the service provider and the user. They do not apply to the wrongs done between users of the service or the platform. The law on this issue is complicated although it may appear deceptively intuitive.

Traditional Choice of Law Rules for Torts

Let’s say you sue the foreign person in a Singapore court; the Court will apply Singapore law as the law of the forum first. The Court will determine that a foreign tort is involved and thus apply a double actionability rule (Parno v SC Marine Pte Ltd [1999] 3 SLR(R) 377 (CA) at [36]; Rickshaw Investments Ltd v Nicolai Baron von Uexkull [2007] 1 SLR(R) 377 (CA) at [56]; JIO Minerals FZC and others v Mineral Enterprises Ltd [2011] 1 SLR 391 (CA) at [88]; Low Tuck Kwong v Sukamto Sia [2014] 1 SLR 639 (CA) at [16]).

The rule is that the alleged wrong must be actionable as a civil wrong in the place where it was committed (lex loci delicti) and also actionable as a tort in Singapore (law of the forum or lex fori).

If that is not the case, then you cannot sue for that wrong in Singapore. You have to sue in that other jurisdiction. However, if that is the case, then the Court will apply Singapore law. The defendant may rely on a defence in either Singapore law or the other law (Low Tuck Kwong v Sukamto Sia [2014] 1 SLR 639 (CA) at [16], which concerned defamation in both Singapore and Indonesia).

The double actionability rule is, however, subject to a “flexible exception” laid down by Lord Wilberforce in Boys v Chaplin [1971] AC 356 (HL) at 391. Under this exception, the tort might nevertheless be actionable in the forum even though one of the two limbs is not satisfied. This makes it possible to have the action determined by a third possible law which has the most significant relationship with the occurrence and the parties. The Court of Appeal in Rickshaw Investments Ltd v Nicolai Baron von Uexkull [2007] 1 SLR(R) 377 (CA) at [58] held that the flexible exception would generally only apply if “the lex fori and/or the lex loci delicti are purely fortuitous and the application of either or both limbs of the ‘double actionability rule’ would result in injustice and unfairness”.

This double actionability rule, although seemingly well-established in Singapore law, has been heavily critiqued to be archaic and ripe for removal (Tan Ming Ren, Revisiting the Double Actionability Rule in Singapore: Time for a Change (2021) 1 SJLS 155; William Tong, Singapore Private International Law On Torts: Inappropriate For Modern Times? [2007] 2 SJLS 405). Further, the rule “would likely deter plaintiffs with foreign tort claims from litigating in Singapore” (IM Skaugen SE v MAN Diesel & Turbo SE [2016] SGHCR 6).

In the Canadian case of v Goldhar [2018] 2 SCR 3, the Canadian Supreme Court had eschewed the double actionability rule altogether for a multijurisdictional defamation case and instead considered two possible alternatives: (i) the law of the place where the most substantial harm to the plaintiff’s reputation occurred; and (ii) lex loci delicti. The significance is the recognition that a universal rule for choice of law for torts, viz. the double actionability rule, is not necessary nor optimal. The lex loci delicti is generally the choice of law rule for tort in China and the European Union (EU) (under the Rome II Regulation). The United Kingdom, just before Brexit and post-Brexit, follows the EU’s Rome II Regulation. Regrettably, the double actionability rule remains well-entrenched in Singapore and the apex court has deemed it fit to retain the law as it is.

The next issue in the legal analysis is tricky. Where was the tort committed? This is needed to determine the lex loci delicti.

The act was done after all in an online virtual world. How does one locate it physically in a territory?

The Court applies the “substance of the tort” test by examining the series of events constituting the elements of the tort to determine where, in substance, the cause of action arose (Wing Hak Man v Bio-Treat Technology Ltd [2009] 1 SLR(R) 446 (HC) at [26]; EFT Holdings, Inc and another v Marinteknik Shipbuilders (S) Pte Ltd and another [2014] 1 SLR 860 (CA) at [53]).

This depends on the tort in question. For negligence, it would depend on whether the allegation of breach of duty of care is regarding manufacture, distribution or sale of the product. For misrepresentation, it is where the representation is received and acted upon unless fortuitous or if the receipt and reliance occur in different countries. If the representation is made generally and not specifically to the claimant, then the place is where the representation is made. For conspiracy, various factors include identity, importance, and location of the conspirators, locations where any agreements or combinations took place, the nature and places of the concerted acts or means, the location of the claimant and the places where the claimant suffered losses. For inducement of breach of contract, various factors include where the inducement took place, where the induced breached occurred, locations of the claimants and defendants, and places where loss was suffered. (See Halsbury’s Laws of Singapore Vol 6(2) (LexisNexis, 2013) at [75.378]).

New Wines in Old Wineskins

The problem with this myriad of connecting factors is that in the online virtual world, the connection to a territory does not really make sense.

This is best illustrated by considering the Australian case of Dow Jones & Company Inc v Gutnick [2002] HCA 56; (2010) 210 CLR 575, which involved defamation. The Court considered two competing theories for identifying the place where the tort of defamation occurred: (i) the place where reputation is harmed i.e. where the defamatory material is read and understood; and (ii) where the material was published.

In an online context, theory (i) would result in the place being where the material is downloaded onto the computer of a person who used a web browser to pull material from the web server. Theory (ii) would result in the place being where the server is maintained. We can immediately see the difficulties with this.

Theory (i) means that a potential defendant could be liable without end in many different jurisdictions. And the law of any of these jurisdictions could apply. This is in a sense pro-claimant. The claimant can shop around for a suitable forum and also an applicable law which is most advantageous to him.

Theory (ii) means that it is fortuitous as to where the server hosting the publication is located. It is based on the legal fiction that a widely disseminated communication is deemed to be a single communication regardless of how many people received it, or where. It is arbitrary that one’s legal rights are possibly determined by where the web server is. This is generally less advantageous to claimants. If the defendant is the internet publisher and if she gets to control the location of the server hosting the publication, then she gets an upper hand.

Either case, it appears intuitively wrong because there’s no actual connection between the real place of the publication, which is the online virtual world accessible by anyone from anywhere in the world on one hand, and on the other hand, the jurisdiction that is selected. We are fitting a wholly different paradigm of the virtual world into an old concept of territoriality. It is new wines in old wineskins.

Of course, there are other legal doctrines which come into play that may narrow the arbitrariness of the above. Doctrines like jurisdiction, forum non conveniens and justiciability may operate ultimately to result in an intuitively sensible outcome as to the forum and applicable law. Nonetheless, the unnecessary complexity of the analysis makes it extremely difficult for potential litigants and even lawyers to navigate what should be a rather simple matter (as usual, it the lawyers who complicate matters.)

Rethinking Choice of Law for the Online Virtual World

Some scholars have proposed various alternative approaches to the doctrine of choice of law in relation to the internet. One approach posits that the Court should simply consider the strongest connection between the parties and any given jurisdiction. Another would be to consider the strongest connection between the dispute and any given jurisdiction. The irony of course is that such an approach does not generate more certainty or clarity. The open-endedness of this approach may be intuitively right but difficult to apply in any given scenario. Yet, it is not unprecedented as well. The Restatement of the Law, Third: Foreign Relations Law of the United States adopts such an approach. In Singapore, the forum non conveniens analysis also considers all possible connecting factors to arrive at a selection of the appropriate forum.

The intuitive simplicity of the approach would at least be sufficient for clarity of application in most cases. While legal disputes in the virtual world remain to be resolved in territory-based jurisdictional courts, this may well be the best solution yet.

Of course, technological and sociological developments may eventually result in a scenario where litigants may bring their disputes in virtual courts. The need for territory-based jurisdictional courts presupposes that legal remedies may be enforced against persons in the jurisdiction by the state’s legitimate sanction and who would otherwise refuse to consent to some other jurisdiction or sanction. However, if the practical remedy to be sought is predominantly virtual or digital, and the means of enforcement of such a remedy would in any event have to be effected through the relevant virtual platform, then arguably the operator of the virtual platform is the equivalent of the state exercising legitimate sanction to enforce the remedy. By analogy, the operator of the virtual platform is justified, and even incentivized, in establishing a dispute resolution scheme which binds the relevant parties, determines the justice of the dispute, and facilitates practical recourse to enact justice.

Increasingly, in the Web 3.0 and blockchain spheres, parties are using smart contracts to implement private dispute resolution schemes which allow disputes to be resolved on-chain and with the practical recourse automated in implementation on-chain. The use of multi-signature addresses, which enable more than one private key to access the relevant cryptoassets, operate a little like an escrow account which allows the unlocking of the asset by a third party adjudicator. There are also crowdsourced blockchain dispute resolution systems which incentivize online jurors or adjudicators to determine the dispute by rewarding them with fees for voting on what turned out to be the “winning party”.

It is not difficult to also imagine a virtual court established in a metaverse platform where the platform operator would be bound to enact the decision of the virtual court through sanctions enforced within the virtual world itself. (The platform could just as well be a blockchain protocol and the equivalent of the operator the protocol’s associated community or Decentralised Autonomous Organisation (DAO).) Will such a virtual court have or develop its own jurisprudence? If the parties are all users of the platform or protocol, they could and would probably be bound by contract to refer their disputes to such a virtual court.

That said, whether such terms of the contract would be necessarily enforceable would ultimately still fail to be determined under the applicable governing law. Traditional legal analysis and doctrines would then apply. Interestingly, a now defunct game and metaverse platform, Second Life, spawned such litigation. In Bragg v. Linden Research, Inc., 487 F.Supp.2d 593 United States District Court for the Eastern District of Pennsylvania, 2007, a user of Second Life sued the operator of the virtual world and its creator for terminating his account. The defendants tried to compel the matter to be referred to arbitration but the Court refused to enforce it as being procedurally and substantively unconscionable.

As such, traditional legal rules remain applicable and relevant. Perhaps in the future, it would be largely secondary to virtual dispute resolution frameworks. Or it may be another string to the bow in a litigant’s selection of possible legal options. Or it may well be that at the end of the futurist dreaming, the dream does not materialise, and we would be back to where we started.

Ronald JJ Wong
Covenant Chambers LLC
E-mail: [email protected]

Ronald JJ Wong engages in both disputes and corporate practices, specialising in technology, intellectual property, corporate finance, financial regulations, and employment.