Tag Archive for: Internet

Conflict of Laws of Freedom of Speech on Elon Musk’s Twitter

/in /by

Elon Musk’s purchase of Twitter has been a divisive event. Commenting on the response on Twitter and elsewhere, Musk tweeted:

The extreme antibody reaction from those who fear free speech says it all

>

By “free speech”, I simply mean that which matches the law.

I am against censorship that goes far beyond the law.

If people want less free speech, they will ask government to pass laws to that effect.

Therefore, going beyond the law is contrary to the will of the people.

Ralf Michaels quote-tweeted perceptively: ‘But which law?’

Twitter and the conflict of laws

By their very nature, digital platforms like Twitter present a variety of conflict of laws issues.

‘Twitter’ is not a monolithic entity. The functionality of the social media platform with which readers would be familiar is underpinned by a transnational corporate group. Twitter, Inc is incorporated in Delaware, and has various subsidiaries around the world; Twitter International Company, for example, is incorporated in Ireland and responsible as data controller for users that live outside of the United States. The business is headquartered in San Francisco but has offices, assets, and thousands of staff around the world.

The platform is populated by 400 million users from all over the world. After the US, the top 5 countries with the most Twitter users are comprised of Japan, India, the UK and Brazil. The tweets and retweets of those users may be seen all over the world. Users have wielded that functionality for all sorts of ends: to report on Russia’s war in real-time; to coordinate an Arab Spring; to rally for an American coup d’état; to share pictures of food, memes, and endless screams; and to share conflict of laws scholarship.

Disputes involving material on Twitter thus naturally include foreign elements. Where disputes crystallise into litigation, a court may be asked to consider what system of law should determine a particular issue. When the issue concerns whether speech is permissible, the answer may be far from simple.

Free speech in the conflict of laws

The treatment of freedom of speech in the conflict of laws depends on the system of private international law one is considering, among other things. (The author is one of those heathens that eschews the globalist understanding of our discipline.)

Alex Mills has written that the balance between free speech and other important interests ‘is at the heart of any democratic political order’.[1] Issues involving free speech may thus engage issues of public policy, or ordre public,[2] as well as constitutional considerations.

From the US perspective, the ‘limits of free speech’ on Twitter is likely to be addressed within the framework of the First Amendment, even where foreign elements are involved. As regards private international law, the Securing the Protection of our Enduring and Established Constitutional Heritage (SPEECH) Act 28 USC 4101- 4105 (‘SPEECH Act’) is demonstrative. It operates in aid of the constitutional right to freedom of expression and provides that a US ‘domestic court shall not recognize or enforce a foreign judgment for defamation unless the domestic court determines that’ the relevant foreign law would provide the same protections for freedom of speech as would be afforded by the US Constitution.[3]

Other common law jurisdictions have approached transnational defamation issues differently, and not with explicit reference to any capital-c constitutional rights. In Australia, the High Court has held that the lex loci delicti choice-of-law rule combined with a multiple publication rule means that defamation is determined by the law of the jurisdiction in which a tweet is ‘available in comprehensible form’: the place or places it is downloaded.[4] In contrast, where a claim concerns a breach of confidence on Twitter, an Australian court is likely to apply the equitable principles of the lex fori even if the information was shared into a foreign jurisdiction without authorisation.[5] In either case, constitutional considerations are sidelined.

The balance to be struck between free speech on the one hand, and so-called ‘personality rights’ on the other, is a controversial issue within a legal system, let alone between legal systems. So for example, the choice-of-law rule for non-contractual obligations provided by the Rome II Regulation does not apply to personality rights, as a consensus could not be reached on point.[6] Similarly, defamation and privacy are excluded from the scope of the HCCH Judgments Convention by Art 2(1)(k)–(l).

There is a diversity of approaches to choice of law for cross-border infringements of personality rights between legal systems.[7] But the ‘law applicable to free speech on Twitter’ is an issue that goes far broader than personality rights. It touches on as many areas of law as there are aspects of human affairs that are affected by the Twitter platform. For example, among other things, the platform may be used to:

Issues falling into different areas of law may be subject to different choice-of-law rules, and different systems of applicable law. What one system characterises as an issue for the proper law of the contract could be treated as an issue for a forum statute in another.

All of this is to say: determining what ‘the law says’ about certain content on Twitter is a far more complex issue than Elon Musk has suggested.

The law applicable to online dignity

Key to the divisiveness of Musk’s acquisition is his position on content moderation. Critics worry that a laissez-faire approach to removing objectionable content on the platform will lead to a resurgence of hate speech.

Musk’s vision for a freer Twitter will be subject to a variety of national laws that seek to protect dignity at the cost of free speech in various ways. For example, in April, the European Parliament agreed on a ‘Digital Services Act’, while in the UK, at the time of writing, an ‘Online Safety Bill’ is in the House of Commons. In Australia, an Online Safety Act was passed in 2021, which provided an ‘existing Online Content Scheme [with] new powers to regulate illegal and restricted content no matter where it’s hosted’. That scheme complements various other national laws, like our Racial Discrimination Act 1975, which outlaws speech that is reasonably likely, in all the circumstances, to offend, insult, humiliate or intimidate another person or a group of people, and was done because of the race, colour or national or ethnic origin of the person or group.

When a person in the United States posts content about an Australian that is permissible under US law, but violates Australian statute, the difficulty of Musk’s position on the limits of censorship becomes clear. Diverse legal systems come to diverse positions on the appropriate balance between allowing online freedom and protecting human dignity, which are often struck with mandatory law. When your platform is frequented by millions of users all over the world, there is no single ‘will of the people’ by which to judge. Perhaps Musk will embrace technological solutions to give effect to national standards on what sort of content must be censored.

A host of other conflicts issues

Musk-era Twitter is likely to pose a smorgasbord of other issues for interrogation by conflict of laws enthusiasts.

For example: legal systems take diverse approaches to the issue of whether a foreign parent company behind a platform like Twitter can be imposed with liability, or even criminal responsibility, for content that is on the platform. While conservatives in America consider the fate of s 230 of the Communications Decency Act—a provision that means that Twitter is not publisher of content they host—other countries take a very different view of the issue. Litigation involving the companies behind Twitter is likely to engage courts’ long-arm jurisdiction.

Perhaps the thorniest conflicts problem that may emerge on Musk’s Twitter is the scope of national laws that concern disinformation. In an announcement on 25 April, Musk stated:

‘Free speech is the bedrock of a functioning democracy, and Twitter is the digital town square where matters vital to the future of humanity are debated’.

Recent years have shown that the future of humanity is not necessarily benefited by free speech on social media. How many lives were lost as a result of vaccine-scepticism exacerbated by the spread of junk science on social media? How many democracies have been undermined by Russian disinformation campaigns on Twitter? The extraterritorial application of forum statutes to deal with these kinds of issues may pose a recurring challenge for Musk’s vision.[8] I look forward to tweeting about it.

Michael Douglas is Senior Lecturer at UWA Law School and a consultant in litigation at Bennett + Co, Perth.

 

[1] Alex Mills, ‘The Law Applicable to Cross-border Defamation on Social Media: Whose Law Governs Free Speech in “Facebookistan”?’ (2015) 7 Journal of Media Law 1, 21.

[2] See, eg, International Covenant on Civil and Political Rights, art 19(3).

[3] SPEECH Act s 3; United States Code, title 28, Part VI, § 4102. See generally Lili Levi, ‘The Problem of Trans-National Libel’ (2012) 60 American Journal of Comparative Law 507.

[4] Dow Jones & Co Inc v Gutnick (2002) 210 CLR 575.

[5] But see Michael Douglas, ‘Characterisation of Breach of Confidence as a Privacy Tort in Private International Law’ (2018) 41 UNSW Law Journal 490.

[6] Art 4(1); see Andrew Dickinson, The Rome II Regulation (Oxford University Press, 2008).

[7] See generally Symeon C Symeonides, Cross-Border Infringement of Personality Rights via the Internet (Brill, 2021) ch VI; Tobias Lutzi, Private International Law Online: Internet and Civil Liability in the EU (Oxford University Press, 2020) ch 4.

[8] See generally Matthias Lehmann, ‘New Challenges of Extraterritoriality: Superposing Laws’ in Franco Ferrari and Diego P Fernández Arroyo (eds), Private International Law: Contemporary Challenges and Continuing Relevance (Edward Elgar, 2019) ch 10.

Worldwide Removal Order Upheld Against Google

/2 Comments/in /by

The Supreme Court of Canada has upheld, by a 7-2 decision, an injunction issued by lower courts in British Columbia requiring Google, a non-party to the litigation, to globally remove or “de-index” the websites of the defendant so that they do not appear in any search results.  This is the first such decision by Canada’s highest court.

In Google Inc. v Equustek Solutions Inc., 2017 SCC 34 (available here) Equustek sued Datalink for various intellectual property violations relating to the manufacture and sale of a networking device.  Interlocutory orders were made against Datalink but it did not comply and it cut any connections it had to British Columbia (para 7).  It continued its conduct, operating from an unknown location and selling its device over the internet.  After some cooperative efforts with Google (de-indexing specific web pages but not Datalink’s entire websites) were unsuccessful to stop potential customers from finding Datalink’s device, Equustek sought an interlocutory injunction stopping Google from including any parts of Datalink websites in its search results worldwide.  Google acknowledged that it could do this relatively easily (paras 43 and 50) but it resisted the injunction.

The issue of the British Columbia court’s in personam or territorial jurisdiction over Google featured prominently in the lower court decisions, especially that of Justice Fenlon for the British Columbia Supreme Court (available here).  This is an interesting issue in its own right, considering the extent to which a corporation can be present or carry on business in a province in a solely virtual (through the internet) manner (rather than having any physical presence).  There is considerable American law on this issue, including the much-discussed decision in Zippo Manufacturing v Zippo Dot Com Inc., 952 F Supp 119 (WD Pa 1997).  In the Supreme Court of Canada, Google barely raised the question of jurisdiction, leading the court to state that it had not challenged the lower courts’ findings of in personam and territorial jurisdiction (para 37).  So more on that issue will have to wait for another case.

The majority decision (written by Abella J) applies the standard three-part test for an interlocutory injunction (para 25).  In doing so it confirms two important points.  First, it holds that a non-party can be made subject to an interlocutory injunction.  It relies on considerable jurisprudence about Norwich orders and Mareva injunctions, both of which frequently bind non-parties.  The common theme the court draws from these cases and applies to this case is the necessity of the non-party being bound for the order to be effective.  In the majority’s view, the injunction against Google is a necessity if the ongoing irreparable harm to Equustek is to be stopped (para 35).  Second, it holds that an interlocutory injunction can be made with extraterritorial effect in cases in which the court has in personam jurisdiction over the entity being enjoined (para 38).  Again, it made such an extraterritorial order in this case because that was, in its view, necessary for the injunction to be effective.  An order limited to searches or websites in Canada would not have addressed the harm.

The dissenting judges (Cote J and Rowe J) accept both of these important points of law.  They acknowledge that the court has the ability, in law, to issue such an injunction (para 55).  But on the facts of this case they determine that the injunction should not have been granted, for several reasons.  First, the injunction is not interlocutory but rather permanent, so that more restraint is warranted.  In their view, Equustek will not continue the action against Datalink, content to have obtained the order against Google (paras 62-63).  In response, the majority notes it is open to Google to apply in future to have the order varied or vacated if the proceedings have not progressed toward trial (para 51).  Because they consider the injunction to be permanent, the dissenting judges object that no violation of Equustek’s rights has as of yet been established on a balance of probabilities (para 66) such that there is no foundation for such a remedy.  Since the majority considers the injunction to be interlocutory this issue does not arise for it.

Second, the dissent rejects the reliance on Norwich orders and Mareva injunctions, noting that in those cases the order does not enforce a plaintiff’s substantive rights (para 72).  In essence, this order is a step farther than the courts have gone in previous cases and not one the dissent is willing to take.  The dissent also denies the injunction because (i) it is mandatory in nature rather than prohibitive, (ii) it is unconvinced that the order would be effective in reducing harm to Equustek and (iii) it thinks there is sufficient evidence that Datalink could be sued in France so that an alternative to enjoining Google is available.  Aspects of this supplementary reasoning are open to debate.  First, the distinction between mandatory and prohibitive orders is not overly rigid and in any event mandatory orders are possible, especially in cases in which the target of the order can easily comply.  Second, common sense suggests the injunction would have at least some impact on the ongoing alleged violations, even though of course there are other internet search engines.  Moreover, the majority points out that it is “common ground that Datalink was unable to carry on business in a commercially viable way unless its websites were in Google’s search results” (para 34).  On the issue of effectiveness, the dissenting judges do not seem to be on this common ground.  Third, proceedings against Datalink in France might or might not be viable.  Even if it could be found in France, it could subsequently leave the jurisdiction and continue its operations elsewhere.  So this seems a hard basis on which to deny Equustek the injunction.

It is fair for the dissent to point out that this injunction is not perfectly analogous to Norwich orders and Mareva injunctions.  It does move beyond those cases.  The debate is whether this is a reasonable incremental move in the jurisprudence relating to the internet or goes too far.  The majority’s overarching rationale for the move is the necessity of the injunction on these facts.  Coupled with the ease with which Google can comply, this is a sufficient basis to evolve the law in the way the court does.