Regulating Internet Hate: A Flying Pig?

The internet is one of the most powerful contemporary tools used by individuals and groups to express ideas and opinions and receive and impart information. [1] It “magnifies the voice and multiplies the information within reach of everyone who has access to it.” [2] Notwithstanding the positive aspects of this development in the realm of free speech and the exchange of ideas, the internet also provides a platform for the promotion and dissemination of hate. [3] In fact, the internet has seen a sharp rise in the number of extreme-right websites and activity. [4] As well as facilitating the promotion of hate, the internet has also strengthened the far-right movement more generally by bringing hate groups together, converging the lines of previous fragmentation, thereby contributing to the creation of a “collective identity that is so important to movement cohesiveness.” [5] This has occurred on an international level, “facilitating a potential global racist subculture.” [6] Although hate existed long before the creation of the internet, this technological advancement has provided an effective and accessible means of communication and expression for hate groups and individuals whilst simultaneously adding a new dimension to the problem of regulating hate, [7] particularly due to the nature of the internet as a global and, to an extent, anonymous medium. It is the anonymity of the internet which deeply hampers the implementation of traditional legal procedures and enforcement of traditional laws, [8] as the perpetrator cannot readily be determined; whilst the global nature of the internet means that, even if a perpetrator can be identified, bringing him or her to justice may not be possible due to jurisdictional limitations. [9] Thus, technological advances in the form of the internet have altered our conceptualisation of a State which habitually had jurisdiction over the activities occurring within its boundaries. To put it simply, this medium knows no borders.

In light of the significant role of the internet vis-à-vis the promotion and dissemination of hate, this paper will look at the issue of regulating the internet in the ambit of hate speech as digitally expressed by individuals and groups. To do so, it will provide a definitional framework of hate speech, an overview of the internet’s role in the ambit of hate speech and consider the challenges in legally regulating online hate speech through a discussion of relevant case-law as well as the Additional Protocol to the Cybercrime Convention. The paper’s analysis emanates from the premise that, if the internet is to be dealt with in a manner which reflects an adherence to principles such as non-discrimination and equality, ”a new template for addressing cross-border contracts” [10] is urgently required. To this end, a comprehensive and unified multijurisdictional approach must be adopted. However, this has proved difficult to date particularly given the stark contrast in the approach vis-à-vis free speech adopted by the United States of America (USA), on the one hand, and Europe on the other. Essentially, as will be reflected hereinafter, it is the conceptual understanding of the scope of the freedom of expression which deeply hampers the creation of an effective regulatory framework for internet hate speech.

Hate speech does not enjoy a universally accepted definition, [11] with most States and institutions adopting their own definitions, [12] notwithstanding that the term is often incorporated in legal, policy, and academic documents. [13] Although non-binding, one of the few documents which has sought to define hate speech is the Recommendation of the Council of Europe Committee of Ministers on hate speech. [14] It states that this term is to be ”understood as covering all forms of expression which spread, incite, promote or justify racial hatred, xenophobia, anti-Semitism or other forms of hatred based on intolerance, including: intolerance expression by aggressive nationalism and ethnocentrism, discrimination and hostility against minorities, migrants and people of immigrant origin.” Interestingly, this definition incorporates the justification of hatred as well as its spreading, incitement and promotion, allowing for a broad spectrum of intentions to fall within its definition. However, it leaves out characteristics such as sexual orientation, gender identity and disability. Hate speech has also been mentioned, but not defined, by the European Court of Human Rights (ECtHR). For example, the Court has refers to hate speech as: “all forms of expression which spread, incite, promote or justify hatred based on intolerance including religious intolerance.” [15] In Vejedland v Sweden, in the framework of homophobic speech, the Court held that it is not necessary for the speech “to directly recommend individuals to commit hateful acts”, [16] since attacks on persons can be committed by “insulting, holding up to ridicule or slandering specific groups of the population” [17] and that speech used in an irresponsible manner may not be worthy of protection. [18] Through this case, the Court drew the correlation between hate speech and the negative effects it can have on its victims, demonstrating that it is not merely an abstract notion, but one with potential to cause harm. In addition, the Fundamental Rights Agency (FRA) of the European Union has offered two separate definitions of hate speech with the first being that it “refers to the incitement and encouragement of hatred, discrimination or hostility towards an individual that is motivated by prejudice against that person because of a particular characteristic.” [19] In its 2009 Report, the FRA held that the term hate speech, as used in the particular section “includes a broader spectrum of verbal acts including disrespectful public discourse.” [20] The problematic part of this definition is the broad reference to disrespectful public discourse, especially since institutions such as the ECtHR extend the freedom of expression to ideas that “shock, offend or disturb.” [21] The Council Framework Decision 2008/913/JHA of 28 November 2008 on Combating Certain Forms and Expressions of Racism and Xenophobia does not directly define hate speech, but instead prohibits different forms of expression and acts that fall within the framework of “Offences Concerning Racism and Xenophobia.” [22] More specifically, Article 1 therein holds that each Member State shall punish the public incitement to violence or hatred directed against a group of persons or a member of a group defined by reference to race, colour, religion, descent or national or ethnic origin, the commission of such an act through public dissemination of material as well as the acts of publicly condoning, denying or grossly trivialising particular crimes such as genocide. This definition could be used in the realm of hate speech but is limited only to particular groups, leaving out others such as sexual minorities. In addition, the threshold of this definition is set to hatred or violence, and does not integrate other “softer” elements of hate speech, such as discrimination. No particular reference to internet hate was made in the above document; however, nothing in its wording prevents it from being used for cases of internet hate. In 2016, the “Code of Conduct on Countering Illegal Hate Speech Online” was signed by different IT companies and the European Commission. This document underlines that the aforementioned Council Framework Decision must be enforced by Member States in online, as well as offline, environments. In the framework of academic commentary, there has been a plethora of definitions put forth to describe hate speech. According to Mari Matsuda, hate speech contains a tripartite definition, namely that the message is “of racial inferiority, the message is directed against historically oppressed groups and the message is persecutory, hateful and degrading.” [23] Wrestling offers a broad interpretation of hate speech including “virtually all racist and related declensions of noxious, identity-assailing expression could be brought within the wide embrace of the term.” [24] Alexander Tsesis has described it as a “societal virus”, [25] while Rodney Smolla refers to the lack of contribution hate speech makes to the development of society since it “cannot contribute to a societal dialogue and therefore can be ethically curtailed.” [26] Scholars, such as Kent Greenawalt have argued about the damaging consequences of such speech, arguing that “epithets and slurs that reflect stereotypes about race, ethnic group, religion and gender may reinforce prejudices and feelings of inferiority in seriously harmful ways.” [27] In discussing bans on racist speech, Post examines several arguments that have been put forth as justifications for such bans including, the “intrinsic harm of racist speech” [28] insofar as there is an “elemental wrongness” [29] to such expression, the infliction of harm to particular groups and individuals, as well as to the marketplace of ideas. [30]

From the above definitions and the variations therein, although some common elements can be discerned, it could be argued that “hate speech seems to be whatever people choose it to mean.” [31] For the purpose of this paper, and taking into consideration that there is no one universal definition of hate speech, a broad definitional basis is embraced. As such, hate speech is hereinafter considered to mean speech that is targeted towards individuals due to their particular characteristics, such as race, ethnic origin, nationality, religion, language, sexual orientation, gender identity and/or disability.

The significant role of the internet in any modern society was recognised by the European Court of Human Rights in Times Newspaper Ltd v UK:

“in light of its accessibility and its capacity to store and communicate vast amounts of information, the Internet plays an important role in enhancing the public’s access to news and facilitating the dissemination of information generally.” [32]

However, as noted above, the internet can also result in harmful expression and this reality began to surface predominantly during the 1990s. In 1994, the UN Secretary General noted that new technologies such as computer programmes, video games and the Minitel system in France were used to disseminate anti-Semitic ideas. [33] In 1995, the UN Special Rapporteur on contemporary forms of Racism, Racial Discrimination, Xenophobia and Related Intolerance recorded the growing use of electronic media for purposes of international communications between far-right groups. [34] In 1996, the UN Secretary General officially recognised the use of the internet and electronic mail as being increasingly used by racist organisations to spread their ideology. [35] In 1997, the aforementioned Rapporteur noted that “the Internet has already captured the imagination of people with a message, including purveyors of hate, racists and anti-Semites.” [36] The first racist website to enter the online world was Stormfront.org set up by a former Ku Klux Klan member and launched in 1995. [37] The dramatic rise of internet hate is reflected by the figures gathered by the Simon Wiesenthal Centre which, in 1995 recorded only one racist website [38] whereas by 2011 its Digital Terrorism and Hate Report had found 14,000 websites, forums and social networks which promoted hate. [39] However, these figures must be considered with a degree of caution, since monitoring becomes more complicated given that websites surface and re-surface at a very fast pace. [40] Websites are not the only sub-tool of the internet with forums, blogs, social networking sites, emails, newsletters, chat rooms and online games being used and abused by extremist groups. Social networking sites have become “breeding grounds for racist and far-right extremist groups to spread their propaganda”, [41] with the sheer number of users, the accessibility to such platforms and the lack of pre-screening of posts or the establishment of, inter alia, Facebook groups, rendering the prospect of regulation a daunting one. In 2008, following a complaint lodged by Martin Shulz, Facebook banned several pages used by Italian extremists to promote violence against Roma. [42] It must be noted that those who spread hate speech may use the internet to harass the victims of their rhetoric directly, to communicate amongst themselves and build up a “sense of belonging and social identity” [43] to a unified movement, but also to recruit new members through the dissemination of their ideology to unsuspecting users who may be confronted with such speech through, amongst others, web links or emails. [44] Further, hate groups attract new members, particularly young people, through the use of innovative methods such as online hate games including “Ethnic Cleaning” and “Shoot the Blacks”.

Thus, the internet which has been named the “network of networks” [45] offers endless possibilities for hate groups to communicate with each other, recruit new members and harass their victims due to its vastness, accessibility and nature as a boundary-free entity governed by no single institution or State. It is the very nature of the internet, and the fact that its effective regulation is contingent upon a common universal approach, which has contributed to its regulation posing a particularly challenging problem for law-makers.

Commencing in the 1990s, several calls were made for more to be done regarding regulating online hate speech. In 1996, the European Commission against Racism and Intolerance (ECRI) requested Council of Europe States to ensure that expression disseminated through the internet which incites discrimination, hate or violence against racial, ethnic, national or religious groups be classed by national law as criminal offences and that such offences should also incorporate the production, dissemination and storage for distribution of harmful material. [46] In 2000, ECRI issued a general policy recommendation on combating the dissemination of racist, xenophobic and anti-Semitic material via the internet, recommending that States ensure that relevant national laws also apply to material uploaded on the internet and to prosecute the perpetrators of relevant offences. [47] ECRI also recommended the clarification of the responsibility of the content host, content provider and site publishers in the framework of the dissemination of racist, xenophobic, and anti-Semitic content over the internet. [48] In 2001, the Declaration and Programme of Action of the Third World Conference against Racism, Racial Discrimination, Xenophobia and Related Intolerance noted that States must “implement legal sanctions, in accordance with relevant international human rights law, in respect of incitement to racial hatred through new information and communication technologies, including the internet.” [49] In 2003, partly as a response to the fact that the ECRI recommendations on internet hate regulation were not adhered to by the Member States, the Council of Europe took the first and only concrete step in seeking to provide a harmonised approach to the regulation of online hate speech, through its Additional Protocol to the Cybercrime Convention. It must be noted that, although there is a general consensus amongst organisations such as the Council of Europe, the United Nations, the European Union and the Organisation for Security and Co-operation in Europe, that internet hate should be regulated, [50] the UN Special Rapporteur on Freedom of Expression held that excessive regulation of the internet in order to “preserve the moral fabric and cultural identity of societies is paternalistic.” [51] However, no extrapolation was made on what could fall within the framework of excessive regulation and, thus, no further conclusions can be drawn thereof.

It is common practice for States to take the position that “what is illegal and punishable in an offline format must also be treated as illegal and punishable online.” [52] However, as the internet is owned by nobody and everybody and knows no physical, electronic, cyber, abstract or other boundaries, it allows its users to transmit messages beyond any such boundaries; thereby, rendering control, censorship and regulation a difficult task.

It is generally accepted that the complexities created by jurisdictional issues constitute the biggest challenge in the sphere of regulating online hate. [53] This is because the internet is not marked by boundaries, cannot be controlled or censored comprehensively by individual States in particular situations and, as a consequence, questions are raised as to “which law should apply and how to delimit competing jurisdictions.” [54] This results in States finding it “difficult to govern and control the flow of information inside and outside their nation states.” [55] More specifically, regulation problems arise where, for purposes of the present discussion, the hateful material is created within a State which does not prohibit the dissemination of xenophobic or racist material on the internet but, due to the boundary-free nature of the internet it is accessible to or, in fact, uploaded by persons residing in a Contracting State. In fact, on a technical level, it is a relatively simple task for individuals who wish to publish information that may be prohibited in some countries, including their own, to go “forum shopping” by choosing internet service providers which are located in countries which permit such content so as to be sure that, notwithstanding potential restrictions in some jurisdictions, the material will be available online. [56] Forum shopping results in the establishment of hate havens, which individuals and groups conveniently choose as hosts for their material. This is the situation in the USA, which is a haven for hate websites. [57] The majority of hate websites are based in the USA and these include those which seek to avoid anti-hate legislation in their own country. [58] Countries such as Spain, have sought to overcome the consequences of such havens by allowing the judiciary to block internet sites that do not adhere to Spanish law. [59] However, this method presupposes the continuous monitoring of internet sites that are in violation of national law, a huge and complex task which cannot possibly be efficiently carried out. Either way, the availability of havens essentially results in the erosion of the weight and role of national laws that seek to restrict online hate as they can be circumvented by carrying out internet activity in, for example, countries which place more emphasis on the freedom of expression rather than on the negative effects of hate. This results in a “de facto extraterritorial application of the laws of some countries known for their robust protection of freedom of expression”, [60] which in turn contributes to the weakening of the more general principle of State sovereignty vis-à-vis the regulation of internet hate.

In Perrin v UK, the ECtHR was confronted with the question of jurisdiction in the sphere of the internet. It interpreted jurisdiction in a broad sense, arguing that the fact that the applicant’s material was uploaded by the applicant on a website operated and legal in the USA did not free the applicant of his responsibilities under UK law which prohibited such material. [61]As such, the Court considered itself to have competence ratione loci regarding material uploaded [62] on the internet, notwithstanding the location in which the material was uploaded. Furthermore, certain countries have also sought to overcome the issue of jurisdiction on a national level. For example, in Germany the Federal Court held that all material uploaded on the world wide web is answerable to German anti-hate legislation regardless of the country in which this material was created, with the only element posing any sort of significance being its accessibility to German internet users. [63]

The variation in approaches of different States to the issue of hateful expression lies at the heart of jurisdictional limitations in the ambit of regulating internet hate. A State’s approach to the issue of restricting forms of expression will be affected by its own “political, moral, cultural, historical and constitutional values” [64] and it is, in fact, this sharp divergence of legal culture in the realm of speech between the USA and Europe which has hindered the efficacy of any regulatory measures and which has rendered the issue of jurisdiction a serious obstacle thereto. As will be reflected in the discussion on the Additional Protocol to the Cybercrime Convention, the variation of approaches between the USA and Europe has limited any formulation of a functioning regulatory framework of online hate simply because the former adheres to an almost absolutist protection of free speech as per the First Amendment, with the latter seeking restrictions for purposes of ensuring that other fundamental rights and freedoms are exercised, such as that of non-discrimination. More specifically, in the USA, hate speech can be proscribed if it constitutes a “true threat”, a test which was developed in the case of Brandenburg v Ohio, and underlines that free speech can only be limited insofar as advocacy of the use of force “is directed to inciting or producing imminent lawless action and is likely to incite or produce such action.” [65] However, the possibility of this test’s application in the sphere of the internet is doubtful, since the internet’s “impersonal contact cannot be seen as readily meeting the true threat requirement of being likely to incite imminent lawless action.” [66] As well as the true threat test, “fighting words” can also be prohibited under US law, a restriction developed in the case of Chaplinsky v New Hampshire. There, the Court held that expression can be restricted if it is made up of fighting words which were deemed to be those which “inflict injury or tend to incite an immediate breach of the peace”. [67] This doctrine was considered within the framework of racist expression in R.A.V. v City of Saint Paul, in which the Supreme Court held that an ordinance which criminalised the placing of symbols, such as a burning cross or a Nazi swastika on a public or private property, was unconstitutional under the First Amendment. [68] The Court held the ordinance unconstitutional because it “imposes special prohibitions on those speakers who express views on the disfavored subjects of race, color, creed, religion or gender” [69] and that St Paul’s objective to “communicate to minority groups that it does not condone the group hatred of bias-motivated speech does not justify selectively silencing speech on the basis of its content.” [70] Thus, given that the Supreme Court was willing to find that the burning of a cross on a black family’s lawn did not fall within the ambit of fighting words and was thus acceptable speech, it seems improbable that the threshold incorporated from the fighting words test could be derived from racist or other hateful speech that takes place online.

Comparatively speaking, freedom of expression in Europe is more readily limited for purposes of preventing not only violence, but also discrimination and hate targeted towards an individual or a group which has a particular characteristic. This is reflected firstly by the fact that the freedom of expression, as protected by Article 10 of the European Convention on Human Rights, is marked by a series of limitation grounds. Subsequently, the ECtHR developed a multi-fold test to be applied in considering whether the freedom of expression should be permitted including: ascertaining whether the limitation has a legitimate aim; is proportional to the aim pursued; is necessary in a democratic society; and is effectuated for purposes of, inter alia, protecting the rights and freedoms of others. To date, the Court’s jurisprudence shows no tolerance for hate speech as reflected in a variety of cases, [71] one of which will be discussed further below. When seeking to restrict hateful expression, the Court usually opts to use the limitation grounds found in Article 10, but has in some situations [72] applied the prohibition of the abuse of rights clause of the ECHR. As noted by the ECtHR, States must “fight against abuses, committed in the exercise of freedom of speech, that openly target democratic values.” [73] Thus, even though this freedom is undoubtedly significant, it must nevertheless coexist harmoniously with other rights and freedoms, with democracy having the duty militantly to protect itself from the abuse of expressive freedom.

Although it is beyond the scope of this paper to carry out an extensive comparative analysis of the stances adopted by the USA and Europe, as an entity in the form of the Council of Europe as well as individual countries which have harmonised their legislation with the European Convention on Human Rights, it is evident that the approaches of the two are oceans away from each other. With such oceans constituting dividing lines, and taking into account the necessity of coherence vis-à -vis online regulation given the nature of the internet, the question of jurisdiction remains a key issue.

The Council of Europe’s Convention on Cybercrime is the first multilateral treaty that aims to combat crimes committed through computer systems and has, to date, been ratified by 47 countries. [97] This Convention was signed and ratified not only by Council of Europe States, but also by the USA which, although is not a member of this entity, has an observer status. Interestingly however, the USA acceded to the Convention only after the issue of online hate was removed from the table of discussions. [98] This reality demonstrates that “fundamental disagreements remain as to the most appropriate and effective strategy for preventing dissemination of racist messages on the Internet”, [99] which subsequently contribute to the weakening or even nullification of regulatory measures that may be adopted by particular States given that internet regulation requires co-operation for both technical and legal reasons as discussed above. To fill the resulting gaps, the Council of Europe subsequently developed the Additional Protocol to the Cybercrime Convention. This has been ratified by 24 countries. [100] The Council of Europe recognised the limitations in implementing a unilateral approach to the issue of online hate in the form of racist or xenophobic hate and, thereby sought to ensure a common set of standards for participating States and promote co-operation amongst them in the criminalisation of relevant acts. [101] This document is seen as a “supplement” [102] to the Convention so as to ensure that the latter’s procedural and substantive provisions encompass racism and xenophobia online. Thus, a series of the Convention’s articles apply mutatis mutandis to the Protocol under consideration including, amongst others, Article 13 on sanctions and measures and Article 22 on jurisdiction.

However, even at first sight, this document comes with several significant limitations which will be discussed hereinafter. Firstly, as demonstrated in its title, this Protocol tackles only racist and xenophobic hate, completely disregarding other forms of hate on grounds including, but not limited to, sexual orientation, gender identity and disability, whilst religion is considered a protected characteristic within the definitional framework set out by Article 2. Thus, there seems to be an unjustified prioritisation of online hate with the Council of Europe almost arbitrarily seeking to regulate the effects of racism and xenophobia online, leaving victims of other types of hate without a respective legal framework.

The Additional Protocol to the Cybercrime Convention defines what is meant by racist and xenophobic material, underlines the measures to be taken at national level in relation to the dissemination of such material, [103] prohibits racist and xenophobic threats and insults professed through computer systems [104] as well as the denial, gross minimisation, approval or justification of genocide or crimes against humanity. [105] The Protocol also renders the intentional aiding and abetting of any of the above a criminal offence. It must be noted that, unlike Article 9 of the Cybercrime Convention which deals with child pornography, the Protocol does not criminalise the possession and procurement of racist and xenophobic material. [106] As noted in the Explanatory Note of the Protocol, in order to amount to an offence, racist and xenophobic material, insults and revisionist rhetoric must occur on a public level, a point which has been incorporated for purposes adhering to Article 8 of the European Convention on Human Rights. [107]

In relation to the acts that are to be deemed offences, it becomes clear that the freedom of expression is “the sacred cow against which the legislation seeks to justify its apparent encroachment for the sake of providing a measure to prohibit cybercrimes motivated by race hate.” [108] To illustrate this, one can turn to Article 3 on the dissemination of racist and xenophobic material through computer systems, with part 1, therein, providing that:

“each party shall adopt such legislative and other measures as may be necessary to establish as criminal offence under its domestic law, when committed intentionally and without right, the following conduct: distributing or otherwise making available, racist and xenophobic material to the public through a computer system.”

However, Part 3 holds that a party may reserve the right not to apply the above paragraph to those cases of discrimination for reasons of upholding free expression. Thus, the Protocol, as an initiative to combat online hate, has been “thwarted through the compromise they have made to concerns about freedom of expression” [109] with much less regard evidently being given to freedoms such as that of non-discrimination. It could thus be argued that the Protocol undermines itself by its approach in that the Council of Europe has given an unequal and unjustifiable emphasis on expression rather than non-discrimination and equality.  [110]

In relation to general limitations that may be imposed on the applicability of Article 3, Part 2 therein, holds that a State may choose not to attach criminal liability to conduct referred to in Part 1 if this does not promote violence or hatred insofar as other effective remedies are available. This is notwithstanding the fact that in the Protocol’s title, reference is made to the criminalisation of racist and xenophobic acts committed through computer systems. Whilst criminalising racist and xenophobic threats has no option to disregard parts of its provisions, Article 5 on racist and xenophobic insults provides that a State has the right not to apply in whole or in part, Part 1 of this Article, which sets out the legislative and other measures that may be adopted to criminalise racist and xenophobic insults. Although no direct reference to free expression is made here as the justifier of such limitation, it could implicitly be assumed that concerns regarding the freedom of expression led to the formulation of the aforementioned reservation available to those who want it. Reserving the right not to apply a particular provision is also incorporated into the denial, gross minimisation, approval, or justification of genocide or crimes against humanity. Many of the States which ratified the Protocol took the opportunity to incorporate reservations. It generally appears that Article 4 on racist and xenophobic threats is the one granted the most protection as it extends to private as well as public communications, unlike the other acts found in the Protocol, while it gives no opt-out possibility as the others do.

The issue of intent is also significant when seeking to appraise the Protocol. This document renders the dissemination of material, threats, insults and revisionist rhetoric offences illegal as well as aiding and abetting the committal of such offences in the event that such acts and/or expressions are effectuated and/or uttered intentionally. This is particularly significant in the realm of the liability of internet service providers who simply constitute the platform through which problematic speech may arise. The Explanatory Report to the Additional Protocol to the Cybercrime Convention holds that the precise meaning of “intentionally” should be interpreted on a national level. [111] However, it did clearly stipulate that it is not sufficient for an internet service provider which simply constitutes the host of the material to be found guilty of any of the Protocol’s offences if the required intent under domestic law did not exist. [112] Thus, on the one hand it does limit the liability of unknowing ISPs but leaves the general conceptualisation of intent unsure and contingent on national positions. However, the Protocol does not regulate or prohibit the finding of permissive intent in the event that an ISP is made aware of racist or xenophobic material or expression and does not take the necessary measures to remove it, thereby, leaving some doors open for finding potential liability in the inaction of ISPs. Such permissive intent is found, for example, in Germany’s Information and Communications Service Act of 1997, which underlines the liability of ISPs in the event that they knew of hateful content, had the ability to block it, but chose not to. [113] Further, in the realm of ISPs, the Protocol remained silent on the very significant question of jurisdiction in the event of a conflict of law between the hosting country and the other. [114] Although for EU countries, the Directive on Electronic Commerce [115] is applicable with Article 3, therein providing that ISPs are governed by the laws of the Member State in which they are established, [116] the situation is not clear in the event that a non-EU country is involved in a particular dispute. [117]

Although the Protocol may contribute to promoting harmonisation regarding agreed upon principles and procedural, technical and legal cooperation amongst States, the Protocol remains problematic. This is the case not only due to its inherent limitations as described above, but also due to the fact that the USA is not part of it. This, in addition to the absence of any form of extradition treaties between the USA and other countries in the sphere of online hate speech, deeply restricts the efficacy of the Protocol’s aims and objectives. Moreover, it may well appear that the Protocol has sought to achieve the lowest possible common denominator, maybe for purposes of maximising ratification. Either way, the aforementioned delimitations may serve as stumbling blocks when seeking to meet the objectives of the Protocol. Furthermore, as well as limitations as a result of an over-emphasis on the freedom of expression, it could be argued that the Protocol constitutes an ineffective base through which online hate can be restricted since it adopts traditional conceptions of State boundaries, State sovereignty on issues such as the freedom of expression mentioned above, and, more generally, treats the issue of online hate as any other issue of traditional means of communication throwing in the concept of international co-operation without effectively and pragmatically considering the challenges of the internet. However, “the Internet is a very different animal from that we are used to, which requires handling in a different way”, [118] but this has not been taken on board.

Hate and hateful expression existed before the creation of the internet and will continue to exist even if tight regulation of online activity were to be achieved. [119] However, the internet has brought about “socio-technological and legal dilemmas that are difficult to handle from a legal point of view”. [120] Moreover, the issue of online hate is moving in new dimensions, with those who disseminate hate speech finding themselves before an array of possibilities to use and abuse the internet for purposes of communication, recruitment and victimisation. However notwithstanding that some case-law has been formulated on a national, transnational and regional level and, even though the Additional Protocol to the Cybercrime Convention has been formulated, the issue of online regulation has not essentially taken any pragmatically significant steps. Firstly, the Protocol itself is lacking as per its scope, as it is arbitrarily limited to racist and xenophobic speech whilst simultaneously limiting its efficacy for purposes of giving particular protection to the freedom of expression. Secondly, the normative US-European divergence of the understanding of free expression has dramatically affected the regulation that Europe seeks to achieve. As noted by one commentator, the global, boundary-free nature of the internet in conjunction with the absolutist approach to expression, as so adopted by the USA, means that “like chasing cockroaches, squashing one does not solve the problem when there are many more waiting behind the walls – or across the border”. [121] More particularly, even if a website is shut down in Germany for example, it may almost immediately pop up again through an American host. At the same time, American courts are not ready to apply any court orders issued in European countries insofar as they are considered to be contrary to the First Amendment. Thus, at the heart of these differences lie fundamental conflicts of legal thought on speech. Interestingly in the case of Yahoo!, the US court recognised that, given that no international treaty or standards were available in the realm of tackling issues on internet speech, the Court is bound by the First Amendment. However following the Yahoo! judgement, the USA finally had the opportunity to be part of such an agreement, however not only opted out of the Additional Protocol to the Cybercrime Convention, but also made its accession to the convention contingent on the exclusion of this theme from the Convention. In brief, there is no intent at the moment on the part of the USA to be part of such an international collaboration in the field of free speech simply because this State’s understanding of free speech does not endorse regulation of hatefulness unless certain high and immediate thresholds, as discussed above, are applied. The result of this approach is that, due to the technical nature of the Internet, the First Amendment has now taken the position as a “default standard for free speech on the Internet” [122] whether other States like it or not. Thus for the moment, it is safe to say that realistic prospects of internet regulation seem unlikely, especially if traditional and purely legal methods are adopted for this purpose.

* Natalie Alkiviadou is a PhD Candidate at the VU Amsterdam and a Lecturer at the University of Central Lancashire Cyprus.