Liability of Internet Intermediaries for Defamatory Speech – An Inquiry into the Concepts of ‘Publication’ and ‘Innocent Dissemination’
by Jan Oster*
NOTICE: THIS WORK MAY BE PROTECTED BY COPYRIGHT
YOU ARE REQUIRED TO READ THE COPYRIGHT NOTICE AT THIS LINK BEFORE YOU READ THE FOLLOWING WORK, THAT IS AVAILABLE SOLELY FOR PRIVATE STUDY, SCHOLARSHIP OR RESEARCH PURSUANT TO 17 U.S.C. SECTION 107 AND 108. IN THE EVENT THAT THE LIBRARY DETERMINES THAT UNLAWFUL COPYING OF THIS WORK HAS OCCURRED, THE LIBRARY HAS THE RIGHT TO BLOCK THE I.P. ADDRESS AT WHICH THE UNLAWFUL COPYING APPEARED TO HAVE OCCURRED. THANK YOU FOR RESPECTING THE RIGHTS OF COPYRIGHT OWNERS.
Ever since the internet became available to the general public, the liability of internet intermediaries has been subject to considerable legislative, judicial, and public scrutiny. Bearing in mind the peculiarities of the tort of defamation and the fundamental importance of freedom of expression and the right to respect for one’s reputation in this context, the principles of liability of internet intermediaries for online defamation is nevertheless far from being fully explored. As Justice Eady pointed out, ‘none of the decisions so far relating to the role of internet service providers ... has definitively established, in general terms, exactly how such entities fit into the traditional framework of common law principles.’1
This paper undertakes a principled and normative approach to the liability of online intermediaries for defamation committed online. By viewing internet intermediary liability for defamatory speech in the broader context of established doctrines of freedom of expression and defamation, the article will contribute to the demystification of ‘cyberlaw’ as a legal discipline. The article argues that the current state of cyberspace-specific legislation, especially the e-commerce Directive2, requires courts to categorise the actions of online intermediaries a priori and thus forces them to apply the same rules to different types of internet intermediaries. However, these intermediaries, such as website operators and search engines, do not have as much in common as it seems at first glance. As a consequence, the article suggests that the categorical approach to internet intermediary liability for defamatory speech should be replaced with a case-by-case approach based on time-tested principles of the law of defamation. The article develops the idea that there is no ‘internet intermediary liability’ as such, but that such liability depends on the concrete function internet intermediaries exercise. Therefore, the paper chooses the legally neutral term ‘internet intermediary’ to encompass all services that constitute and operate on the internet, such as internet service providers (ISPs), ‘information society service providers’ in EU parlance, website operators, portals, platforms and search engines.
II. The privileges for internet intermediaries under the EU e-commerce Directive
According to the EU’s Internal Market competence, the e-commerce Directive only encompasses services within the meaning of Article 56 of the Treaty on the Functioning of the European Union (TFEU), that is, services normally provided for remuneration.3 As Recital 18 of the Directive explains, ‘information society service providers’ extend to services which are not remunerated by those who receive them, such as advertisers.4 Commercial ISPs will usually fall under this definition.5
The UK implemented Article 12 et seq. of the Directive through Regulations 17 et seq. of the Electronic Commerce (EC Directive) Regulations 2002 (SI 2002 No 2013). For the relationship between the EC Directive Regulations 2002 and the innocent dissemination-based defences in Section 10 Defamation Act 2013 and Section 1 Defamation Act 1996, the principle lex specialis derogat legi generali applies: The EC Directive Regulations 2002 governing the specific subject matter ‘liability of information society service providers’ override the Defamation Acts that cover innocent dissemination defences for defamation in general.6 Such an interpretation is also required by the supremacy of EU law. Where internet intermediaries do not satisfy the definition of an ‘information society service provider’, the ordinary statutory and common law principles of defamation law apply.7
1. General principles
The general rule for the liability of internet intermediaries, as far as they only act as intermediaries and not as content providers, is to be found in Article 15 e-commerce Directive. Accordingly, internet intermediaries are neither under a general obligation to monitor content they transmit or store, nor under a general obligation actively to seek facts or circumstances indicating illegal activity. However, Member States are not prevented from imposing a monitoring obligation on service providers only with respect to obligations of a general nature. This does not concern monitoring obligations in a specific case and, in particular, it does not affect orders by national authorities in accordance with national legislation.8 These specific cases are outlined in Articles 12(3), 13(2) and 14(3) of the ecommerce Directive: The exemption of internet intermediaries from liability granted in Articles 12 to 14 shall not affect the possibility for a court or administrative to require the service provider to prevent or to terminate an infringement. In addition, Article 14(3) enables Member States to establish procedures governing the removal or disabling of access to information. As a consequence, Member States’ authorities can only be empowered to issue injunctions, but cannot award damages against internet intermediaries, if the requirements of the exemptions from liability are fulfilled (see infra 3. and 4.).9 Member States have to ensure that court actions available under their national law concerning information society services’ activities allow for the rapid adoption of measures designed to terminate any alleged infringement and to prevent any further impairment of the interests involved (Article 18(1) ecommerce Directive).
According to Article 15(2) e-commerce Directive, Member States may establish obligations for information society service providers promptly to inform the competent public authorities of alleged illegal activities undertaken by their customers. Yet these obligations do not extend to duties under private law, including defamation, but are restricted to public and criminal law.10 In the UK, ISPs are required to monitor transmissions under the Regulation of Investigatory Powers Act 2000 for the purposes of crime prevention and national security.
2. In particular: no obligation to install filtering software
The Court of Justice of the European Union (CJEU) had to decide whether internet intermediaries could be required to install filtering software in order to prevent an infringement of another person’s intellectual property rights, or whether such an obligation constitutes an obligation to generally monitor data contravening Article 15 e-commerce Directive. The CJEU cases involved SABAM, a Belgian management company which represents authors, composers and publishers of musical works, the access provider Scarlet Extended, and the social networking site Netlog, respectively. The Court categorised Scarlet Extended as a mere conduit and Netlog as a host provider.11 SABAM brought interlocutory proceedings against Scarlet and Netlog, claiming that these companies were best placed, as ISPs, to take measures to end copyright infringements committed by their customers. SABAM requested the installation of a filtering system concerning all electronic communications passing via Scarlet’s services and all information which is stored on Netlog’s servers by its service users, as a preventive measure, exclusively at the ISPs’ expense, and for an unlimited period. The filtering system should be capable of identifying electronic files (Netlog) containing certain copyright-protected works, or the movement thereof (Scarlet), with a view to blocking the transfer of files the sharing of which would be in breach of copyright.
The Court observed that an implementation of such a filtering system would require ISPs to identify, within all of the electronic communications of all their customers, the files which are likely to contain works in respect of which holders of intellectual-property rights claim to hold rights, to establish the lawfulness of these files, and to prevent files that they consider to be unlawful from being made available.12 Preventive monitoring of this kind, the Court held, would require ‘active observation’ of all electronic communications conducted on the network of the access provider and of files stored by users with the hosting service provider concerned.13 This, the Court concluded, amounts to ‘general monitoring’, which Article 15(1) of the e-commerce Directive prohibits.14
3. Articles 12 to 14
Articles 12 to 14 of the e-commerce Directive provide immunity rules applicable to specific services of internet intermediaries. Articles 12 to 14 are based on two principles of accountability: internet intermediaries may face liability for the exercise of editorial control, or because of the intermediary’s intentional or negligent omission to prevent dissemination. According to Recital 42 of the Directive, the exemptions from liability established in the ecommerce Directive ‘cover only cases where the activity of the information society service provider is … of a mere technical, automatic and passive nature, which implies that the information society service provider has neither knowledge of nor control over the information which is transmitted or stored.’
a) Mere transmission of, or provision of access to, information, Article 12(1)
Where the intermediary service consists of the mere transmission of information or the mere provision of access to a communication network (‘mere conduit’), Article 12(1) stipulates that the service provider is not liable for the information transmitted, on the condition that he does not initiate the transmission, does not select the receiver of the transmission, and does not select or modify the information contained in the transmission.
Article 12 aims at transient messages, which pass through the intermediary defendant’s system for purposes of access and which the intermediary neither stores for significant periods, nor modifies.15 The e-commerce Directive indicates in Recital 43 that ‘manipulations of a technical nature which take place in the course of the transmission as they do not alter the integrity of the information contained in the transmission’ do not constitute a modification of the content.16 The requirements for liability stipulated in Article 12—the exceptions from immunity—are justified: Initiating the transmission, selecting the receiver and, in particular, selecting or modifying content goes beyond the role of a merely passive transmitter of third-party speech and constitutes the exercise of a form of editorial control.17 Furthermore, a service provider ‘who deliberately collaborates with one of the recipients of his service in order to undertake illegal acts goes beyond the activities of “mere conduit” ... and as a result cannot benefit from the liability exemptions established for these activities.’18 However, the defendant’s knowledge alone is irrelevant, so he is not liable for failing to take steps to prevent access to another site which he is aware carries defamatory material.19
b) Temporary storage of information by ‘conduits’ (‘caching’), Articles 12(2) and 13
The act of temporarily storing information by internet intermediaries bears hardly any significance for the law of defamation, as the storage of information itself does not constitute a publication and therefore does not give rise to an action in defamation. The main purpose for providing rules on temporary storage of information is the protection of Intellectual Property rights, since the temporary storage of information may already constitute an interference with copyright.20 According to Article 12(2), if the acts of transmission and of provision of access ‘include the automatic, intermediate and transient storage of the information transmitted in so far as this takes place for the sole purpose of carrying out the transmission in the communication network, and provided that the information is not stored for any period longer than is reasonably necessary for the transmission’, the internet intermediary benefits from the privilege according to Article 12(1). Yet, if the network or service provider stores the information for a longer period of time and for other purposes than those provided by Article 12(2), the requirements for an exception from liability, which are stipulated in Article 13(1) e-commerce Directive, are stricter. The purpose of Article 13 is to protect internet intermediaries in respect of material for which they are not the primary host but which they store temporarily on their computer systems for the purpose of enabling the efficient availability of internet material.21 Many internet intermediaries temporarily store commonly accessed web pages on their computer systems so that those pages will be more quickly accessible to their subscribers.22 This ‘caching’ is described as a ‘sort of half way house between mere transmission and “hosting”’.23
c) Permanent storage of information (‘hosting’), Article 14
In contrast to temporary ‘caching’ for technical purposes, permanent storage of information may give cause for liability in defamation. Article 14(1) stipulates that where the information society service consists of the storage of information (‘hosting’), the host provider is not liable for the information stored at the request of the content provider, on condition that he:
(a) does not have actual knowledge of illegal activity or information and, as regards claims for damages, is not aware of facts or circumstances from which the illegal activity or information is apparent (emphasis added); or
(b) upon obtaining such knowledge or awareness, acts expeditiously to remove or to disable access to the information.
The privilege contained in Article 14(1) shall not apply when the author using the service is acting under the authority or the control of the provider (Article 14(2)). In this case, the internet intermediary would be fully accountable for the information like a content provider. Article 14 is based on the fact that host providers are, apart from the content providers themselves, closest to the risk of liability for defamatory content, because they have immediate access to the content. If they have knowledge of illegal activity or information, or if they are aware of facts or circumstances from which the illegal activity or information is apparent, then they have to remove or to disable access to the information concerned, otherwise they may be subject to liability for defamation.
Article 14 e-commerce Directive raises several problems. First, it is ambiguous: it is unclear what kind of services the term ‘host provider’ actually encompasses with regard to the user generated content services postdating the e-commerce Directive. Article 14 has been applied to a wide range of internet intermediaries, including operators of interactive websites24, blog-publishing services25, online chatrooms26, online marketplaces27 and social networks28—types of internet services which are all based on rather different business models. Moreover, the application of the e-commerce immunity legislation to more recent forms of internet services, such as online encyclopaedias29, news aggregators30, hyperlinks and search engines31, is far from clear. Secondly, it is uncertain what ‘expeditiously’ (Article 14(1)(b)) actually means; whether it allows enough time, for instance, to consult a lawyer.32 Thirdly, and most significantly, Article 14 does not provide any guidance regarding the notion of ‘awareness’ where an order to pay damages is at issue. Article 14 requires awareness of facts or circumstances from which the illegal activity or information is apparent. Therefore, as regards claims for damages in defamation, the seminal question is: Does ‘awareness’ according to Article 14 e-commerce Directive only relate to the fact that the allegedly defamatory information is being hosted, is legal knowledge of their defamatory character required, or is even legal knowledge of the non-availability of defences necessary? If the first is correct, then the immunity of internet intermediaries providing host services would be exceptionally narrow. They would have to conduct their own legal research following any notification they receive, because if they remove content that turns out to be lawful, then they might be in breach of their contractual relations with their customers. If it is the second or the third, then the immunity of intermediaries providing host services would depend on their legal knowledge, which in turn means that they could escape liability if they conduct no further research.
By comparison, Regulation 19 of the EC Directive Regulations 2002, which serves to implement Article 14, stipulates that the intermediary must have been ‘aware of facts or circumstances from which it would have been apparent to the service provider that the activity or information was unlawful’. In contrast, Section 1(1)(c) Defamation Act 1996 requires that the defendant knows of his contribution to the publication of a defamatory statement. The difference is that a statement is defamatory even though there is a defence available, whereas ‘illegality’ or unlawfulness includes that the statement is defamatory and there is no valid defence.
The CJEU considered Article 14 e-commerce Directive in the 2011 case of L’Oréal SA and ors v eBay International AG and ors. The case dealt with the question of whether eBay would have a defence under Article 14 in a case of trademark infringement where sellers used eBay to sell counterfeit goods. The Court held that ‘it is sufficient, in order for the provider of an information society service to be denied entitlement to the exemption from liability provided for in Article 14 …, for it to have been aware of facts or circumstances on the basis of which a diligent economic operator should have identified the illegality in question and acted in accordance with Article 14(1)(b) …’.33 The rules set out in Article 14(1)(a) ‘must be interpreted as covering every situation in which the provider concerned becomes aware, in one way or another, of such facts or circumstances.’34 The Court then went on to explain: ‘The situations thus covered include, in particular, that in which the operator of an online marketplace uncovers, as the result of an investigation undertaken on its own initiative, an illegal activity or illegal information, as well as a situation in which the operator is notified of the existence of such an activity or such information. In the second case, although such a notification admittedly cannot automatically preclude the exemption from liability provided for in Article 14 of [the e-commerce Directive], given that notifications of allegedly illegal activities or information may turn out to be insufficiently precise or inadequately substantiated, the fact remains that such notification represents, as a general rule, a factor of which the national court must take account when determining, in the light of the information so transmitted to the operator, whether the latter was actually aware of facts or circumstances on the basis of which a diligent economic operator should have identified the illegality’35 (emphasis added). As a consequence, the Luxembourg Court requires that the ‘awareness’ of the intermediary should be determined by recourse to a set of factors, one of which is apparently the notification of the unlawful content itself. With hindsight, one could argue that the Court indicated this approach already in the 2010 Google France decision: there the Court held that ‘it is necessary to examine whether the role played by that service provider is neutral, in the sense that its conduct is merely technical, automatic and passive, pointing to a lack of knowledge or control of the data which it stores’ (emphasis added).36 The use of the term ‘pointing to’ suggests that merely technical, automatic and passive activity is but one factor to determine whether the intermediary had actual knowledge. However, apart from these two decisions, the Court leaves open what the other factors to determine ‘awareness’ actually are.
III. Plea for an incremental case-by-case approach instead of legislation
Compared with the incremental and inductive approach of case-law, the main advantage of legislation as a form of lawmaking is supposed to be that it provides a higher degree of legal certainty. With regard to the e-commerce Directive in particular, the divergences between the legal situations of the EU Member States led to calls for a special harmonising regime. Because of the EU’s limited area of competences, the adoption of the e-commerce Directive was, in the first place, based on economic considerations. The then-EC recognised that ‘disparities in Member States’ legislation and case-law concerning liability of service providers acting as intermediaries prevent the smooth functioning of the internal market, in particular by impairing the development of cross-border services and producing distortions of competition’.37 The e-commerce Directive should therefore ‘constitute the appropriate basis for the development of rapid and reliable procedures for removing and disabling access to illegal information.’38
With regard to the law of defamation, the e-commerce Directive—and the same considerations apply to any other legislation specifically tailored to liability of particular internet intermediaries, such as Section 5 of the Defamation Act 2013—fails to fulfil its purpose of providing legal certainty. Legislation will inevitably lag behind new developments in cyberspace. The e-commerce legislation from 2000 does not cover recent forms of internet services, such as news aggregators and search engines. It is therefore left to the courts to develop doctrinally sustainable approaches to deal with such providers.39 A recent decision of the German Federal Court of Justice on a defamation claim against a search engine’s autocomplete function reveals the legal uncertainty caused by legislation that does not keep pace with technological developments.40 Furthermore, Articles 12-14 equally apply to several forms of internet intermediaries which do not have as much in common as it seems at first glance (see supra II.3.c)). Eventually, as already shown, the legal requirements of Article 14—knowledge of the illegal publication, or awareness of the facts thereof—are far from being clear. Even an update of the e-commerce Directive would soon be antiquated, as will be Section 5 of the Defamation Act 2013, because further internet services that cannot be conceived of as yet will necessarily follow.
It is therefore argued that the legislative approach to liability of internet intermediaries for defamatory content should be abandoned. Any attempt to categorise internet intermediaries or their actions a priori is futile. Instead, courts should decide on the liability of an internet intermediary on a case-by-case approach, based on ‘straightforward common law principles’ of the tort of defamation41 and with a view to their particular role concerning the publication of an impugned statement.42 This may vary from case to case, depending on the business model of the intermediary rather than an abstract legal categorisation. In order to harmonise the legal situations in the different Member States, courts should enter into a judicial dialogue by taking into consideration and referring to other courts’ decisions, with the European Court of Human Rights as the court of last resort. The Strasbourg Court has already decided its first internet-related cases43, including a case on liability of an internet intermediary44, and further applications concerning liability of internet intermediaries are currently pending45. If there is any phenomenon where such judicial dialogue is needed and expedient, it is the ubiquitous cyberspace. The next paragraph of this article elaborates on this, and explains how the traditional principles of the law of defamation and freedom of expression are suitable and sufficient to cope with the challenges of online intermediary liability.
IV. Internet intermediaries in the dimensions of defamation
Defamation is a tort committed through human communication. The tort of defamation requires, first, a ‘message’: the defamatory statement. Whether a statement is defamatory or not—online as well as offline—has been subject to a considerable amount of case-law46 and now statutory intervention47, and will not be subject to further examination here. Secondly, every defamatory message needs an author or creator. This is also the person who is usually, liable for defamation, but not necessarily: he may be able to invoke a defence deriving from his right to freedom of expression, such as qualified privilege or honest comment, but also based on other principles, such as Parliamentary privilege.48 Thirdly, the author has to share the defamatory message with one or more recipients. Under the common law, the tort of defamation requires that at least one recipient is not identical with the allegedly defamed person, as the tort of defamation protects a person’s reputation and not just a person’s honour. Fourthly, in order to communicate the message, the author or creator has to use a particular medium. Where the author or creator uses the medium ‘spoken words’, the tort committed is slander, whereas defamatory statements issued in permanent form constitute libel. The choice of the communication medium may involve other individuals or companies in the publication process. A slander committed via telephone, for instance, requires a telephone company to establish and hold a connection. A libellous newspaper article needs not only a journalist writing the article, but also a printer to print the article and a newspaper vendor to sell the newspaper. Persons involved in the mere dissemination of a statement of which they are not the author may usually invoke defences when sued for defamation based on the principle of ‘innocent dissemination’, which will be dealt with later. These persons are ‘intermediaries’, the term referring to their role as a ‘medium’ or as the provider of the medium in the communication process.
Despite its paramount contribution to the possibilities of human communication, including the increased opportunities to defame someone, the internet has, as such, not changed the basic dimensions of communication and defamation: the taxonomy of author, message, medium and recipient. Internet communication might complicate the distinction between these components, as will be elaborated further below in this article, but it does not alter them. As a consequence, neither has the internet changed the tort of defamation as a legal concept for a communication tort. The tort of defamation applies to internet communication as well as to other forms of communication. The internet has raised new questions regarding particular requirements of defamation; for instance, whether a sloppy comment in a ‘chat conflict of laws.49 However, a need for a new ‘tort of online defamation’ has not yet occurred.
With a view to the scope of this article, the seminal question is therefore: How do online intermediaries fit into the general scheme of the tort of defamation? The answer depends on whether online intermediaries are either to be perceived as authors or creators of defamatory messages, or whether they merely provide the communication medium as intermediaries. Where the former is the case, they are fully liable for the statement, unless they may rely on a (usually content-based) defence, such as justification, qualified privilege or honest comment. Where the latter is the case, a further question needs to be raised as to the availability of a defence based on the principle of innocent dissemination (see infra 2.). In any case, as will be argued in the next paragraph, online intermediaries who contribute to the distribution of defamatory messages are to be considered ‘publishers’ under defamation doctrine.
1. Speech, intermediaries, and ‘publication’
It is a fundamental tenet of the human rights doctrine that thoughts alone cannot lawfully be subjected to any restriction.50 As a consequence, defamation requires publication of a message to a recipient, in order to be legally comprehensible. Liability in defamation is not restricted to authors of defamatory statements. The gist of a claim in defamation is publication, not creation.51 Thus both the author of a defamatory statement and the person or institution assisting in the dissemination of the statement—the intermediary—can be treated as ‘publisher’, and hence, can be liable for defamation. Whether one distinguishes them into ‘primary’ and ‘secondary’ publishers52 does not alter the fact that they are both publishers. Consequently, any person who takes part in making the defamatory statement known to another may be liable for it.53 Under the common law of England and Wales, publication does not even require active participation; the mere failure to prevent distribution with the defendant’s means is sufficient to be considered a publisher.54 Such speech intermediaries may be, for instance, providers of a notice board, booksellers, libraries, or newspaper vendors.55
It is controversial, however, whether ‘merely passive’ distributors, such as telephone companies or postal services, may indeed be considered ‘publishers’. With regard to the internet, this question is of significant importance with regard to access providers or host providers who are not aware of defamatory content being transmitted through or posted on their facilities.
US courts require that the defendant must have had ‘a direct hand’ in the publication.56 In Cubby v CompuServe, an action was brought against a computer service company that provided its subscribers with access to an electronic library of allegedly libellous news publications put together by independent third parties and loaded onto the company’s computers. The District Court for the Southern District of New York held that the defendant computer service company was a mere ‘distributor’ of information, which could not be held liable for defamatory statements made in news publications, unless the claimant could prove that the defendant company knew, or had reason to know, of defamation. Such distributors, the Court held, are protected under the First Amendment to the US Constitution.57 Similarly, in Lunney v Prodigy, the New York Court of Appeals held that Prodigy, an internet chat room provider, was not considered a publisher of defamatory material posted from an imposter account due to its passive role in monitoring the chat rooms.58
Yet it is questionable whether this doctrine can, or should, be applied as a general rule or at least as a rule applicable to internet defamation cases. Section 230(c)(1) of the US Communications Decency Act 199659, which was enacted after the above-mentioned CompuServe decision, stipulates that ‘[n]o provider or user of an interactive computer service shall be treated as the publisher …’. Courts have interpreted Section 230(c)(1) as providing ISPs total immunity from defamation actions, even where the claimant has given notice to the ISP.60 However, as a matter of statutory construction, it is doubtful whether this provision removes internet intermediaries from the notion of ‘publication’. The provision does not stipulate that an intermediary is not the publisher. Rather, the law creates the legal fiction that a person shall not be treated as the publisher. Thus Section 230(c)(1) indirectly confirms that a provider of an interactive computer service may, in law, be a publisher.
Under the law of England and Wales, it is not entirely clear whether knowledge of the words complained of is relevant for the characterisation as ‘publisher’. Two lines of judicial reasoning can be distinguished.
First, in the early internet-related case Godfrey v Demon Internet Ltd61, Justice Morland held that ISPs, even if they were ignorant of the defamatory material they disseminated, were nonetheless publishers. The case concerned an action in defamation against an ISP that carried a newsgroup called ‘soc.culture.thai’, which would be available to be read by the customers. An unknown person in the United States posted a comment that was considered ‘squalid, obscene and defamatory of the plaintiff’.62 The judge stated: ‘In my judgment the Defendants, whenever they transmit and whenever there is transmitted from the storage of their news server a defamatory posting, publish that posting to any subscriber to their ISP who accesses the newsgroup containing that posting. Thus everytime one of the Defendants’ customers accesses “soc culture thai” and sees that posting defamatory of the Plaintiff there is a publication to that customer.’63 Morland J compared the case with the similar US decision Lunney v Prodigy, where the New York Court of Appeals held that Prodigy was not responsible in law for the e-mail or the defamatory bulletin board postings, one of the reasons being that ‘Prodigy did not publish the statement’.64 Morland J took the view this was an approach which was inconsistent with English law. Instead, Morland J was of opinion that ‘at English common law Prodigy would clearly have been the publisher’ of the message.65
Secondly, in contrast to Morland J’s opinion, in the 2006 judgement Bunt v Tilley, Eady J held: ‘I have little doubt, however, that to impose legal responsibility upon anyone under the common law for the publication of words it is essential to demonstrate a degree of awareness or at least an assumption of general responsibility, such as has long been recognised in the context of editorial responsibility.’66 He concluded that ‘an ISP which performs no more than a passive role in facilitating postings on the internet cannot be deemed to be a publisher at common law.’67 As a consequence, ‘persons who truly fulfil no more than the role of a passive medium for communication cannot be characterised as publishers: thus they do not need a defence.’68 Eady J later upheld the Bunt v Tilley rule in Tamiz v Google, according to which an ISP involved only in dissemination is not to be treated as a publisher unless he knew, or should by the exercise of reasonable care have known, that the publication was likely to be defamatory.69 This reasoning was confirmed by the Court of Appeal.70
Both lines of argument rely on essentially the same precedences, but read them in different ways. Both Bunt v Tilley (at para. 23) and Godfrey v Demon Internet (at para. 31) refer to the 1885 case Emmens v Pottle, where the defendants ran a newsstand which sold, among other papers, a magazine containing defamatory material.71 Lord Esher MR held that it is not enough that a person merely plays a passive instrumental role in the process: ‘I am inclined to think that this called upon the defendants to shew some circumstances which absolved them from liability, not by way of privilege, but facts which shew that they did not publish the libel.’72 This seems to indicate that Lord Esher MR intended to negate any publication on behalf of the defendants. However, his Lordship continued:
‘[T]he defendants did not compose the libel on the plaintiff, they did not write it or print it; they only disseminated that which contained the libel. The question is whether, as such disseminators, they published the libel? If they had known what was in the papers, whether they were paid for circulating it or not, they would have published the libel, and would have been liable for so doing. That, I think, cannot be doubted. But here, upon the findings of the jury, we must take it that the defendants did not know that the paper contained a libel. I am not prepared to say that it would be sufficient for them to shew that they did not know of the particular libel. But the findings of the jury make it clear that the defendants did not publish the libel. Taking the view of the jury to be right, that the defendants did not know that the paper was likely to contain a libel, and, still more, that they ought not to have known this, which must mean, that they ought not to have known it, having used reasonable care – the case is reduced to this, that the defendants were innocent disseminators of a thing which they were not bound to know was likely to contain a libel’ (emphasis added).
As a consequence, Emmens v Pottle can be read both ways: the absence of knowledge of a defamatory statement may already exclude the existence of a publication on behalf of the defendant, or it may constitute a mere defence.
In Bunt v Tilley (at para. 22), Eady J also referred to the 1899 decision McLeod v St. Aubyn as support for his argument. In McLeod, a case which involved accusations of contempt of court, the alleged ‘publication’ consisted of handing over an unread copy of a newspaper for return the following morning, the newspaper containing statements which were allegedly in contempt of court. Lord Morris indeed held, as quoted by Eady J, that a ‘printer and publisher intends to publish, and so intending cannot plead as a justification that he did not know the contents. The appellant in this case never intended to publish.’73 However, there is no evidence to suggest that Lord Morris therefore denied publication of the newspaper by the appellant. His Lordship did not indicate that the lack of intent to publish excludes publication as such. Rather, someone who intends to publish ‘cannot plead as a justification that he did not know the contents’ (emphasis added). Read in context, it becomes clear that Lord Morris was more concerned with the question of legal accountability in general, than with whether there has been a publication on behalf of the defendant. He observed: ‘It would be extraordinary if every person, who innocently handed over a newspaper or lent one to a friend with no knowledge of its containing anything objectionable, could be thereby constructively but necessarily guilty of a contempt of a Court because the said newspaper happened to contain scandalous matter reflecting on the Court.’74 Yet Lord Morris remains unclear whether the question of ‘innocence’ and ‘guilt’ referred to the publication as such, or rather to available defences.
Eventually, both the Court of Appeal in Tamiz v Google (at para. 26) and Morland J in Godfrey v Demon Internet (at para. 34) cite the case Vizetelly v Mudie’s Select Library Ltd75 as support for their argument. Vizetelly concerned the liability of a circulating library for distributing books containing libels. Smith LJ held: ‘The defendants having lent and sold copies of the book containing that libel, primâ [sic] facie they published it. What defence, then, have they? None, unless they can bring themselves within the doctrine of Emmens v. Pottle. ... It seems to me that out of the mouth of Mr. Mudie there was sufficient evidence to justify the jury in coming to the conclusion that the defendants had failed to prove their defence, and that it was through negligence on their part that they did not find out that the book contained a libel on the plaintiff. That being so, they failed to do what the defendants in Emmens v. Pottle succeeded in doing, namely, prove that they did not publish the libel.’76 One can therefore also read Vizetelly as referring to innocent dissemination as a defence or as excluding the requirement of publication.
The practical importance as to whether one perceives lack of awareness as excluding publication as such, on the one hand, or as merely providing a defence, on the other hand, lies in the burden of proof. If it is part of the notion of ‘publication’, then the claimant would have to prove it. If it is a defence, the onus would be on the defendant to prove that he did not know that he distributed defamatory material. Against this background, the precedences seem to indicate that lack of awareness is a mere defence, as they all require the defendant to prove his innocence. With regard to internet intermediaries, placing the burden of proof concerning the lack of awareness and mere passiveness on the defendant appears justified indeed. Whoever provides social internet networks, blogging services or telecommunication infrastructure provides a means of mass communication and draws the benefit from it, and thus, albeit involuntarily, increases the risk of defamation. Against this argument, in Tamiz v Google Eady J compared an internet intermediary transmitting defamatory content with the owner of a wall that has been smeared with defamatory graffiti. Eady J observed that the fact that the owner could acquire scaffolding and have it all deleted with whitewash ‘is not necessarily to say, however, that the unfortunate owner must, unless and until this has been accomplished, be classified as a publisher.’77 However, this comparison is inapt. The difference between an ISP and the owner of a wall is that it is intrinsically the function of an ISP to transmit or store information, parts of which may be, as a matter of fact, defamatory. This does not apply to a house wall.
Moreover, an exclusion of ‘merely passive’ distributors of information from the notion of publication raises difficulties in those cases where the distributer actually knows that their infrastructures are being used for the publication of defamatory material.78 It seems uncontroversial that an internet intermediary—or any other speech intermediary—is to be considered a ‘publisher’ within the meaning of the law of defamation after he has been notified that he transmits or stores defamatory content.79 The authority for this rule is the 1937 Court of Appeal decision Byrne v Deane.80 This case concerned an allegedly defamatory message which someone had posted on the notice board of a golf club. The defendants were the proprietors of the golf club, and one of them was also the club secretary. They had not been involved in the initial posting on the wall, but they failed to remove the statement after they had been notified. Greer LJ explained: ‘In my judgment the two proprietors of this establishment by allowing the defamatory statement, if it be defamatory, to rest upon their wall and not to remove it, with the knowledge that they must have had that by not removing it it would be read by people to whom it would convey such meaning as it had, were taking part in the publication of it.’81 There is no convincing reason not to treat a passive distributor as publisher before he takes notice. The concept of ‘publication’ should not be confused with authorship or endorsement of a statement and it does not decide on a person’s accountability for it. ‘Publication’ does not contain any indication as to its unlawfulness; it is a question of fact.82 Within the system of the law of torts, the concept of publication stands for the general tortious requirement of causation in fact.
Consequently, it is submitted that online intermediaries may be considered ‘publishers’ within the meaning of the law of defamation, if they participate in making the defamatory statement known to another person, disregarding their knowledge of the message. It is argued that the common law of England and Wales contradicts any qualification of the concept of ‘publisher’ in internet-related defamation cases because of passiveness, lack of awareness or insignificance of the internet intermediary’s role. These aspects should be treated as questions of the availability of a defence, but not of publication.83 Publication is not a fault-based concept. As the next paragraph shows, the defence of ‘innocent dissemination’ proves to be the more refined concept to determine accountability in such cases. Therefore, Justice Morland’s reasoning should be followed.
2. The principle of innocent dissemination
The crucial test within the system of the tort of defamation to determine the liability of internet intermediaries, like of any other speech intermediary, is the search for available defences. Intermediaries participate in the publication of the defamatory statement, but they may invoke defences based on the principle of an ‘innocent dissemination’. As the term ‘innocent’ suggests, the principle of innocent dissemination is a principle of legal accountability. As such, innocent dissemination stands in contrast to the factual concept of publication. In the system of the law of torts, the defence of ‘innocent dissemination’ thus resembles the requirement of causation in law. The lawfulness of one’s action or omission is to be decided on the availability or non-availability of an innocent dissemination defence, not upon the publication as such.
The principle of innocent dissemination long predates the internet. It emerged under common law and has been codified by legislative intervention in Section 10 Defamation Act 2013 and Section 1 Defamation Act 1996.84 Some commentators argue that Section 1 Defamation Act 1996 abolished the common law defence of innocent dissemination.85 However, the Defamation Act 1996 does not provide Parliament’s clear and unambiguous intention in this respect86, as was the case, for instance, with the explicit abolition of the Reynolds defence in Section 4(6) Defamation Act 2013 .
The common law defence of innocent dissemination implies that no one may be liable for unlawful content that they merely distribute, but of which they are not the author, otherwise endorse and of which they do not have knowledge and the absence of knowledge was not due to any negligence on their part.88
The principle of innocent dissemination consists of two parts: First, the defendant’s action or omission constituted a mere dissemination rather than a creation of, or association with, a certain statement. Secondly, the dissemination had to be made ‘innocently’, that is, without knowledge of the unlawful content.
a) Mere dissemination
The first question for liability of internet intermediaries is whether they provide their own content or approve of third-party statements, on the one hand, or whether they act as mere speech intermediaries, on the other hand. Section 10(1) of the Defamation Act 2013 provides that a court ‘does not have jurisdiction to hear and determine an action for defamation brought against a person who was not the author, editor or publisher of the statement complained of unless the court is satisfied that it is not reasonably practicable for an action to be brought against the author, editor or publisher.’ For a definition of ‘author’, ‘editor’ and ‘publisher’, Section 10(2) refers to Section 1 of the Defamation Act 1996. Accordingly, ‘”author” means the originator of the statement, but does not include a person who did not intend that his statement be published at all; “editor” means a person having editorial or equivalent responsibility for the content of the statement or the decision to publish it; and “publisher” means a commercial publisher, that is, a person whose business is issuing material to the public, or a section of the public, who issues material containing the statement in the course of that business.’89 Section 10 Defamation Act 2013 and Section 1 of the Defamation Act 1996 grant a defence90 to distributors who are ‘publishers’ within the meaning of the law of defamation, because they took part in the publication of the statement. By comparison, Section 230(c)(1) US Communications Decency Act formulates that ‘[n]o provider or user of an interactive computer service shall be treated as the publisher…’ (emphasis added), thus acknowledging that a person is a publisher in a legal sense, but should not be treated as one.
The question of whether online intermediaries act as content providers or mere distributors of third-party content is not susceptible to a general definition, but depends on the particular case. The argument that a distinction between the provision of content and the mere dissemination thereof may be difficult to establish in internet-related cases does not relieve the legal practitioner from this undertaking. Line-drawing problems are an essential part of legal work, and so far the courts have been fairly successful in deciding such cases despite the vague guidance of the e-commerce Directive.91 For instance, while the e-commerce Directive eschewed a statement on liability for hyperlinks and merely made it subject to a report on the application of the Directive (see Article 21(2)), the Supreme Court of Canada has already decided that hyperlinks to a particular website may under certain circumstances be regarded as the own publication of the hyperlinker.92 Furthermore, the German Federal Court of Justice recently decided that terms suggested by a search engine’s autocomplete function93 are the search engine operator’s own speech and not merely the reflection of the entries of search engine users. Eventually, the High Court decided that an internet intermediary goes beyond mere dissemination of information if he exercises editorial control over the information concerned.94
If the internet intermediary acted as a mere intermediary and did not provide own content or did not associate himself with third-party content, the second question is whether the dissemination occurred innocently. In contrast to the concept of publication, innocent dissemination is fault-based. The defendant has a valid defence if he did not know that he was distributing unlawful material, and that this ignorance was not due to negligence.95 In contrast to the statutory defence in Section 1 Defamation Act 1996, the common law defence of innocent dissemination also exculpates a disseminator who knows that the statement is defamatory, but had reason to believe that the author could invoke a valid defence, such as justification, qualified privilege or honest comment.96
The question of whether an online intermediary acted innocently, it is argued, is not susceptible to a generalising answer; rather, the concrete function exercised by the internet intermediary, their access to impugned content and their knowledge of the unlawful information or the negligence concerning their lack of knowledge needs to be scrutinised on a case-by-case basis. The decisive factors establishing liability of internet intermediaries for defamatory content have to be found by balancing the conflicting rights and interests of the persons involved. These are freedom to impart information on behalf of the content provider and the corollary right of the public to receive information (Article 10 ECHR or, within the ambit of EU law97, Article 11 EUChFR) and the intermediary’s freedom to conduct a business (Article 16 EUChFR)98 as well as his right to impart information under Article 10 ECHR and Article 11(1) EUChFR, on the one hand, and the right to respect for one’s reputation as a sub-category of the right to respect for private life according to Article 8 ECHR and Article 7 EUChFR99, on the other hand. Notably, the Luxembourg Court has not even mentioned internet intermediaries’ right to impart information so far.
Neither of these conflicting principles takes precedence per se.100 Rather, they have to be balanced against each other on a case-by-case basis. The instrument that provides rational conditions for this balancing exercise is the principle of proportionality. The proportionality test examines:
(1.) whether the interference with the rights of the internet intermediary was suitable to achieve the legitimate aim pursued, that is, the protection of the claimant’s reputation (suitability);
(2.) whether the interference was the least intrusive instrument amongst those which might achieve the legitimate aim (necessity); and
(3.) whether the interference was strictly proportionate to the legitimate aim pursued (proportionality strictu sensu).101
An internet intermediary’s liability is not ‘necessary’ if there was another measure available to pursue the aim that would have been less restrictive but nonetheless equally effective. A less intrusive means a court would have to consider before holding an internet intermediary liable for defamation is the liability of the content provider, who is fully liable for the statements he authored. The gravity of the conflict therefore lies between the allegedly defamed claimant and the content-provider, and not between the claimant and the internet intermediary, who acts as a mere messenger. Internet intermediary liability should therefore be subsidiary vis-à-vis content provider liability. In order to establish the identity of an anonymous content provider, a court may have to issue a Norwich Pharmacal order102 in accordance with the users’ right to protection of their personal data.103 Especially internet intermediaries providing host services should also contractually reserve the right to identify their customers in order to ensure that there is a right to an effective remedy against persons who have used their services to publish defamatory content. It is also fair that the inferior party of these main proceedings has to bear the internet intermediary’s cost, as far as they are reasonable.104
Another factor in a court’s judgment should be the subsidiarity of an internet intermediary to pay damages for a transmitted or hosted defamatory statement vis-à-vis an injunction requiring the blocking of access or removal of the impugned information. It is submitted that an internet intermediary should only be required to remove or block access to allegedly defamatory material if a court requests the intermediary to do so. An intermediary should not be required to remove content upon its own initiative or upon request by a victim of defamatory material (‘notice and take down’), because such a regulatory environment creates an incentive for the intermediary to take down content from the internet that is higher than the potential costs of not taking it down.105 Intermediaries should not be made the legal gatekeeper of the internet. Yet in order to avoid liability towards their customers, internet intermediaries should reserve a take-down right in their contracts.
A service provider who is simply notified of a defamatory allegation, in circumstances where it is not clear whether a defence is available, will not have actual knowledge of unlawful information or awareness of facts or circumstances from which it would have been apparent to the service provider that the information was unlawful. First, it is often difficult to establish whether the meaning was defamatory or not. Secondly, even positive knowledge of the defamatory meaning alone should not suffice to hold the intermediary liable; the information may be true, or the content provider may have other defences available the intermediary is not aware of. In Tamiz v Google, Eady J held: ‘Accordingly, a provider who simply receives notification that particular words are alleged to be defamatory will not have received notification of illegality in terms that are adequately substantiated. Such a provider would not have actual knowledge of illegality; nor an awareness of facts or circumstances from which it would have been apparent that the information was unlawful. In order to achieve that state of mind, it would be necessary to examine and consider, on an informed basis, the validity or strength of any available defences (including, for example, those of justification, fair comment and qualified privilege in one or other of its recognised forms).’106 In addition, the blocking of internet content is technically only feasible through interferences with data protection rules107—a question the CJEU has not even dealt with yet—and may even constitute a criminal offence under Section 1 of the Regulation of Investigatory Powers Act 2000.