Assignment Essay

“Fake news”: reconsidering the value of untruthful expression in the face of regulatory uncertainty Irini Katsirea

Centre for Freedom of the Media, Department of Journalism Studies, University of Sheffield, Sheffield, UK

ABSTRACT Against the backdrop of the regulatory furore over ‘fake news’, this article examines the protection that is afforded to untruthful expression by the European Court of Human Rights and by national courts in Germany, the UK and the US. It argues that the suppression of ‘fake news’ in the face of uncertainty over the contours of this highly politicised term and of the evidentiary vacuum as to the harm posed, may run counter to constitutional guarantees of free speech. Regulatory interventions seeking to curb the flow of ‘fake news’, which is not per se illegal, require careful consideration lest they should empower governments or unaccountable technology corporations without editorial culture to become the arbiters of truth.

KEYWORDS Fake news; freedom of expression; Germany; United Kingdom; United States

Introduction

The phenomenon of ‘fake news’, which came to the fore as a result of Buzz- feed’s revelations about the commercial exploitation of fabricated news stories on Facebook, and allegations of Russian interference in the US, French and German elections by way of propaganda campaigns, has gained increased cur- rency in recent times and sparked fears over the threat posed to democracy.1

In response to this perceived threat, EU Member States are in the process of drawing their strategy in this undeclared media war. Germany adopted a new law to give teeth to its existing criminal law sanctions by imposing heavy fines against social media platforms that fail to speedily delete fake reports and hate speech.2 The UK has been more cautious in its endeavour to come to grips with this phenomenon. Differently from Germany, there has been no

© 2019 Informa UK Limited, trading as Taylor & Francis Group

CONTACT Irini Katsirea [email protected] This article has been republished with minor changes. These changes do not impact the academic content of the article. 1C. Silvermann and D. Alexander, ‘How Teens in the Balkans are Duping Trump Supporters with Fake News’ <https://www.buzzfeed.com/craigsilverman/how-macedonia-became-a-global-hub-for-pro-trump- misinfo?utm_term=.abwQX0Y5JL#.psmLE4WAZp>.

2For other European initiatives such as the French ‘fake news’ bills of October 2018, see LSE Commission on Truth, Trust and Technology, ‘Tackling the Information Crisis: A Policy Framework for Media System Resi- lience’, November 2018, 51 <http://www.lse.ac.uk/law/news/2018/truth-trust-technology>.

JOURNAL OF MEDIA LAW 2018, VOL. 10, NO. 2, 159–188 https://doi.org/10.1080/17577632.2019.1573569

legislative initiative with the aim of combatting ‘fake news’ in the UK so far. The House of Commons Digital, Culture, Media and Sport Committee has invited submissions to a Fake News Inquiry, which have fed into an interim report, while a more substantial report is expected in 2019. Mean- while, social media platforms and search engines, in an attempt to keep stric- ter regulation at bay, have responded by cooperating with fact-checking organisations and by reducing the financial incentives for the production of ‘fake news’ content. It is a moot point whether such initiatives are capable of tackling the challenge of ‘fake news’. Some argue that the ‘fake news’ phenomenon is intrinsic to social media’s and, in particular, Facebook’s business model, so that they have little interest in meaningfully addressing it.3 Others are more hesitant to discredit social media’s attempts at self-regu- lation.4 They agree though with well-established research on the so-called ‘illusory truth effect’, which suggests that fact-checking is likely to further entrench erroneously held beliefs rather than eradicate them.5

The other side of the coin is the troubling use of the term ‘fake news’ by the US President but also by nationalist, far-right parties such as the German parties Alternative for Germany (AfD) and Patriotic Europeans against the Islamisation of the West (Pegida) for political advantage. However, the term ‘fake news’, translated into German as ‘Lügenpresse’, as well as the fears associated therewith and the perception that their spread needs to be put to a halt, have a long history. The Trump administration and nationalist parties who lambast the mainstream media in their tweets, election campaigns and demonstrations join a long tradition of press victimisation. In the First World War, the notion of ‘Lügenpresse’ was enlisted in the effort to discredit reporting by the enemy. Before the NS party’s seizure of power, this concept was weaponised against the ‘unpatriotic’ press of the Weimar Republic, which failed to stand up to the demeaning Versailles Treaty; later it was used against foreign media, not least by the chief Nazi propagandist Joseph Goebbels.6

These eerie comparisons suggest that the misuse of the term ‘fake news’ is per- ennial, and that it should be resisted together with all attempts to stifle pur- veyors of misinformation. The history of this concept as a tool to crack down on dissent is but one reason for scepticism. The vagueness, highly

3P. Bernal, ‘Fakebook: Why Facebook Makes the Fake News Problem Inevitable’ (2018) 69 (4) Northern Ireland Legal Quarterly 513.

4B. Holznagel, ‘Phänomen “Fake News” – Was ist zu tun? Ausmaß und Durchschlagskraft von Desinforma- tionskampagnen’ (2018) 1 MultiMedia und Recht 18.

5G. Pennycook, T. D. Cannon and D.G. Rand, ‘Prior Exposure Increases Perceived Accuracy of Fake News’ (2018) 147(12) Journal of Experimental Psychology 1865; P. Schneiders, ‘Gegen Fake News ist niemand immun’, <http://www.ard.de/home/ard/Was_die_Wissenschaft_zu_Fake_News_sagt/3733254/index. html?articleSectionIndex=0>.

6R. Blasius, ‘Unwort des Jahres: Von der Journaille zur Lügenpresse’ Frankfurter Allgemeine Zeitung (Frankfurt am Main, 13 January 2015) <http://www.faz.net/aktuell/gesellschaft/unwort-des-jahres-eine-kleine-geschichte- der-luegenpresse-13367848.html>; see T. McGonagle, ‘“Fake News”: False Fears or Real Concerns?’ (2017) 35(4) Netherlands Quarterly of Human Rights 203, 205 et seq for historic examples of ‘fake news’.

160 I. KATSIREA

politicised nature and possible inadequacy of the term ‘fake news’ to capture the many facets of our complex information ecosystem might be another.7

This article will begin by exploring the meaning of ‘fake news’ so as to establish a working definition that would avoid the pitfalls of its recent distor- tions. The Culture, Media and Sport Committee’s interim report recommends that the term be rejected and replaced by a shared definition of the terms ‘mis- information’ and ‘disinformation’.8 While these terms are less politically loaded than the term ‘fake news’, one needs to pay heed to the fact that the term ‘fake news’ is likely here to stay as ‘part of the vernacular that helps people express their frustration with the media environment’.9 It is therefore useful to explore its contours further. Next, this article will discuss whether ‘fake news’ pose a threat that would justify their regulation. The answer to this question is by no means clear given that regulatory and legal solutions are debated largely in a vacuum of evidence as to the necessity of their adop- tion. Lastly, our attention will turn to the single most important reason why our efforts to sanitise our news ecosystem might be ill-conceived. This would be the case if the spreading of untruthful but not illegal information, even of such that is outright fabricated, with intent to deceive, was protected under the right to freedom of expression. It needs to be borne in mind that ‘fake news’ are not automatically illegal if they do not violate laws on privacy, defa- mation, hate speech, misleading advertising etc. This section will discuss, first, the verification obligations that the European Court of Human Rights imposes on the media and the extent to which such obligations extend to new media actors. Secondly, it will consider the level of protection, which should be afforded to ‘fake news’ in the hierarchy of expression recognised by the Court. In other words, do ‘fake news’ still deserve to be protected as political speech or do they constitute quasi-commercial expression that could be regulated by laws akin to those applicable to misleading advertising? Thirdly, it will be asked whether fruitful conclusions can be drawn from the Court’s jurisprudence in cases of historical revisionism as regards the possi- bility of regulating ‘fake news’ by way of the criminal law ‘armoury’. The final section of this article will consider the protection afforded to untruthful expression in the German, UK and US legal systems. The diverse

7C. Wardle and H. Derakhshan, ‘Information Disorder: Toward an Interdisciplinary Framework for Research and Policymaking’, Council of Europe report DGI (2017) 09, 27 September 2017 <https://rm.coe.int/ information-disorder-toward-an-interdisciplinary-framework-for-researc/168076277c>; European Association for Viewers Interests, ‘Infographic: Beyond Fake News – Ten Types of Misleading News – Nine Languages’ <https://eavi.eu/beyond-fake-news-10-types-misleading-info/>.

8House of Commons Digital, Culture, Media and Sport Committee, ‘Disinformation and “Fake News”: Interim Report’, HC 363, 29 July 2018, <https://publications.parliament.uk/pa/cm201719/cmselect/ cmcumeds/363/363.pdf>.

9R. K. Nielsen and L. Graves, ‘“News You Don’t Believe”: Audience Perspectives on Fake News’ (Oxford Uni- versity, Reuters Institute for the Study of Journalism, Factsheet October 2017) <https://reutersinstitute. politics.ox.ac.uk/sites/default/files/2017-10/Nielsen%26Graves_factsheet_1710v3_FINAL_download. pdf>.

JOURNAL OF MEDIA LAW 161

constitutional traditions of these jurisdictions, informed by different historical pasts, shape the extent to which they are prepared to protect false statements of fact and make for an interesting comparison.

What is ‘fake news’?

Craig Silverman, the Buzzfeed author who traced back the source of hoax stories to a few enterprising teenagers in the small town of Veles, Macedonia, defined ‘fake news’ as ‘100-per-cent-false stories predominantly published by sites that exclusively traffic in hoaxes to generate clicks’.10 A precise definition of the term ‘fake news’ is essential if one is to be able to distinguish this phenomenon from other forms of inaccurate information as well as from truthful news. One might think that this term could be a useful shorthand so as to distinguish between legitimate news outlets and unreliable ones, in other words so as to separate the wheat from the chaff. However, the other side of the coin is President Trump’s use of the term ‘fake news’ as a weapon against traditional media, which in turn throw the accusation at each other or fight back as in the case of the US newspapers’ campaign in August 2018.11 It is important to make a distinction between: knowingly false content circulated by the media, such as where the error has been due to carelessness; other stories, which contain mistakes, even though they have been researched with due care; and finally, stories that are not outright false but are exaggerated, biased or tendentious.

All programmes broadcast in the UK, including those of the BBC, are required to comply with obligations of due accuracy and due impartiality laid down in s. 5.1 of the Ofcom Broadcasting Code.12 The BBC Editorial Guidelines explain that accuracy

‘is not simply a matter of getting facts right. If an issue is controversial, relevant opinions as well as facts may need to be considered. When necessary, all the relevant facts and information should also be weighed to get at the truth’.13

The BBC has been criticised time and again for getting this balance wrong by giving undue weight to minority scientific viewpoints, for example on the

10G. Cunningham, ‘Avoid the Misnomer “Fake News”’, 20 February 2017 <https://villagemagazine.ie/index. php/2017/02/avoid-the-misnomer-fake-news/>.

11D. Nuccitelli, ‘The Mail’s Censure Shows Which Media Outlets Are Biased on Climate Change’, 25 Septem- ber 2017 <https://www.theguardian.com/environment/climate-consensus-97-per-cent/2017/sep/25/ the-mails-censure-shows-which-media-outlets-are-biased-on-climate-change>; ‘US Media Fight Back Against Trump Attacks’, 16 August 2018 <https://www.dw.com/en/us-media-fight-back-against- trump-attacks/a-45101206>.

12Ofcom, ‘The Ofcom Broadcasting Code (with the Cross-Promotion Code and the On Demand Programme Service Rules)’, 3 April 2017, <https://www.ofcom.org.uk/tv-radio-and-on-demand/broadcast-codes/ broadcast-code>; An Agreement Between Her Majesty’s Secretary of State for Culture Media and Sport and the British Broadcasting Corporation (Cm 9366), December 2016, Schedule 3.3 <https://www.bbc. co.uk/corporate2/insidethebbc/managementstructure/bbccharterandagreement>.

13BBC Editorial Guidelines, <https://www.bbc.co.uk/editorialguidelines/guidelines>.

162 I. KATSIREA

MMR vaccine or on climate change.14 It has even been argued that imparti- ality rules consign broadcasters to a ‘mealy-mouthed neutrality giving the same prominence to a false as to a true statement’.15 This should not be the case. Impartiality does not mean giving equal weight to all sides of a debate without weighing the evidence. At the same time, it is necessary to bear in mind that today’s scientific heresy might become tomorrow’s orthodoxy. An interpretation of impartiality that seeks to broaden the range of voices to be heard rather than silence unpopular ones is much needed in our polarised world.16

As far as the print media are concerned, both the Editors’ Code of Practice followed by the Independent Press Standards Organisation (IPSO), as well as the Independent Monitor for the Press (IMPRESS) Standards Code, in craft since 24 July 2017, stipulate in their very first clause that their respective member publishers need to adhere to the obligation of accuracy.17 Since January 2016, the IPSO Code has been supplemented by a new reference to the need for headlines to be supported by the text of the article beneath.18

An example of a story that was found to contravene this provision was a story published by The Sun on 9 March 2016 under the headline ‘Queen backs Brexit’. IPSO determined that this story was significantly misleading given that it contained an unsupported allegation that the Queen had funda- mentally breached her constitutional obligation to remain strictly neutral as regards political matters.19 The text beneath claimed that the Queen made critical comments about the EU at a lunch at Windsor Castle in 2011 but did not refer to the Queen’s position on Brexit. Is this an example of ‘fake news’? In order to answer this question, it is necessary to explore the term ‘fake news’ further.

Silverman’s is but one of a number of recent attempts at capturing the meaning of the elusive term ‘fake news’. IMPRESS, in its submission to the Department for Culture, Media and Sport (DCMS) ‘Fake News’ Inquiry, defined ‘fake news’ as ‘the knowing and consistent publication of

14K. Marsh, ‘Kevin Marsh, ex-Executive Director, BBC College of Journalism on Issues of Impartiality in News and Current Affairs’ (2012) 1 (1) Journal of Applied Journalism and Media Studies, 69, 76; for a recent case of a breach of the obligation of due accuracy under s. 5.1 of the Ofcom Code see Ofcom’s Broadcast and On Demand Bulletin, Issue 351 of 9 April 2018, ‘BBC Radio 4, 10 August 2017, 6:00’, p. 12.

15M. Doherty, ‘Should Making False Statements in a Referendum Campaign Be an Electoral Offence?’, 4 July 2016, <https://uk.constitutionallaw.org>.

16R. Sambrook, ‘Delivering Trust: Impartiality and Objectivity in the Digital Age’, July 2012, <https:// reutersinstitute.politics.ox.ac.uk/our-research/delivering-trust-impartiality-and-objectivity-digital-age>; see, however, the BBC’s position on climate change reporting in L. Hickman, ‘Exclusive: BBC issues internal guidance on how to report climate change’, 7 September 2018, <https://www.carbonbrief. org/exclusive-bbc-issues-internal-guidance-on-how-to-report-climate-change>.

17IPSO, ‘Editors’ Code of Practice’ <https://www.ipso.co.uk/editors-code-of-practice/> (referred to in the following as ‘IPSO Code’); IMPRESS, ‘The IMPRESS Standards Code’ <https://impress.press/standards/ impress-standards-code.html>.

18IPSO Code, Clause 1 s. 1. 19IPSO, Decision of the Complaints Committee 01584-16 Buckingham Palace v The Sun, 20 April 2016 <https://www.ipso.co.uk/rulings-and-resolution-statements/ruling/?id=01584-16>.

JOURNAL OF MEDIA LAW 163

predominantly false information in the guise of news’.20 These definitions seek to exclude close cousins of ‘fake news’ such as: inadvertent errors included in news publications; tendentious and misleading but not outright false news stories; conspiracy theories which are hard to prove to be true or false, but originate from people who believe them to be true; inaccurate state- ments not ‘in the guise of news’, for instance such that are published in sites that do not hold themselves out to be authoritative news organisations or such that go hand in hand with strategies of persuasion in political speeches.21 If one was to apply these definitions to the ‘Queen Backs Brexit’ story mentioned above, one would come to the conclusion that it does not constitute ‘fake news’. One may argue about The Sun’s quality, but it would be an exagger- ation to say that it exclusively or even consistently publishes false information. If one was to tar all media that publish biased, tendentious, or even occasion- ally – to a greater or lesser extent – inaccurate stories with the brush of ‘fake news’, this would discredit them, and would further undermine public trust in the media and the ability of the latter to act as ‘public-watchdogs’.22 Also, if one was to label mainstream media as ‘fake news’, this would beg the question as to what is the touchstone against which to measure media reliability, the antipode of ‘fake news’.

‘Fake news’ is but one permutation of many different types of potentially misleading content in our information eco-system. Clare Wardle of First Draft, a non-profit organisation seeking to improve truth and trust online, drafted a typology of seven forms of mis- and disinformation. Among them feature satire or parody with no intention to cause harm but having the poten- tial to fool; misleading use of information to frame an issue or individual; imposter content that impersonates genuine sources; fabricated content that is 100% false and is designed to deceive and do harm; false connection when headlines, visuals or captions do not support the content; false context when false content is shared with false contextual information and, finally, manipulated content when genuine information or imagery is manipulated to deceive. These types of ‘problematic content’ sit on a scale according to the degree of the intent to deceive.23

Indeed, intent to deceive is key when trying to draw the line between ‘fake news’ and other forms of misleading information. Let us take the example of news satire. News satire is a genre that is very popular with the young and that

20IMPRESS, ‘Written Evidence Submitted by IMPRESS: The Independent Monitor for the Press’, March 2017 <https://www.parliament.uk/business/committees/committees-a-z/commons-select/culture-media- and-sport-committee/inquiries/parliament-2015/inquiry2/>.

21H. Allcott and M. Gentzkow, ‘Social Media and Fake News in the 2016 Election’ (2017) (31) (2) Journal of Economic Perspectives 211, 214.

22IMPRESS, Trust in journalism sinks to all-time low as YouGov Poll reveals public demand for decent stan- dards of journalism, 5 December 2016 <http://www.impress.press/news/yougov-poll.html>.

23C. Wardle, ‘Fake News. It’s Complicated’, 16 February 2017 <https://firstdraftnews.com/fake-news- complicated/>.

164 I. KATSIREA

might have the potential to re-invigorate their political interest. Satirical pub- lications ordinarily aim to mock, not to deceive. The problem is that such publications are often mistaken for real. The editor of The Onion, an estab- lished satire news organisation considers that they have missed their target when this happens.24 Such confusion is unlikely to arise when they make plain their intent to satirise, say by way of a disclaimer. The implication could be that satirical content, which does not make the satirical intent obvious, and has hence the potential to mislead, constitutes ‘fake news’. Indeed, many articles published on satirical websites are often mistaken for real, especially when they are viewed on isolated Facebook or Twitter feeds. A story that was published on the now defunct satirical website wto5news.- com that Pope Francis had endorsed Donald Trump’s presidential candidacy was shared over one million times on Facebook. Many people bought into it despite the fact that the ‘About’ section of the website disclosed its nature, albeit without placing a disclaimer in the individual stories. There is a fine line between ‘fake news’ and ‘satire’, and tarring them with the same brush risks casting doubt over legitimate forms of expression, and further aggravates the crisis of trust over online communication.25 One could try to draw the line by characterising only such satirical content as ‘fake news’ that is disseminated by way of isolated Twitter or Facebook feeds without a disclaimer. This would, however, mean that one would need to treat the same story differently depending on its presentation and dissemination, and that the most delicate forms of satire would be hit the hardest. It seems preferable to place emphasis on the intent to deceive and to take the overall character of the site in question into account.26 Checking the URL of a site, examining the source of a report and looking for other reports on the same topic are some of the standard prac- tices by way of which to spot ‘fake news’.27 Having said that, the extent to which satirical content is legally protected speech, depends on intricate, fact-specific factors, and can only be determined on a case-by-case basis.28

We have established so far that ‘intent to deceive’ is a key criterion, dis- tinguishing ‘fake news’ from other, more innocent forms of misleading infor- mation such as news satire. As became apparent from the abovementioned

24A. Meade, ‘The Onion in the Age of Trump: ‘What We Do Becomes Essential when its Targets are this Clownish’ <https://www.theguardian.com/culture/2017/aug/28/the-onion-in-the-age-of-trump-what- we-do-becomes-essential-when-its-targets-are-this-clownish>.

25D. Coast, J. Fox and D. Welch, ‘Written Evidence’, March 2017 <https://www.parliament.uk/business/ committees/committees-a-z/commons-select/culture-media-and-sport-committee/inquiries/ parliament-2015/inquiry2/>; see Vereinigung Bildender Künstler v Austria (2008) 47 EHRR 5, para 33.

26Cf Bladet Tromsø and Stensaas v Norway, Application Nummer 21980/93 at [63]. 27A. Mosseri, ‘A New Educational Tool Against Misinformation’, 6 April 2017 <https://newsroom.fb.com/ news/2017/04/a-new-educational-tool-against-misinformation/>; L. Bounegru, J. Gray, T. Venturini and M. Mauri, ‘A Field Guide to Fake News. A Collection of Recipes for those Who Love to Cook with Digital Methods’, January 2018 <https://fakenews.publicdatalab.org/>.

28Nazi slut case, Landgericht (Regional Court) Hamburg, Az. 324 O 217/17, 11 Mai 2017; Hustler Magazine, Inc. v Falwell, 485 U.S. 46, 56 (1988).

JOURNAL OF MEDIA LAW 165

typology, ‘fake news’ is not a homogeneous concept, but can manifest itself in different ways. In an effort to shed more light on the diverse forms of infor- mation disorder that plagues our news ecosystem, the Council of Europe dis- tinguished in 2017 between three types of deceptive or otherwise harmful content: misinformation, which is false information, not created with the intent of causing harm; disinformation, which is false information, which is deliberately created to cause harm; and mal-information, defined as infor- mation based on reality, but shared to inflict harm, often by moving it from the private to the public sphere.

The last of these three types of information disorder, namely mal-infor- mation, is beyond the scope of this article that focuses on content not grounded in reality. The only feature distinguishing misinformation from dis- information is the agent’s motivation to do harm. The Council of Europe report argues that this motivation ‘provides a deeper understanding of how dis- or mal-information campaigns work, it also points to possible ways to resist them’.29

While the question of intent to do harm is relevant in criminal law terms, it does not enhance definitional clarity. It is not a criterion commonly found in political science literature, and begs the question as to which types of intended outcome might amount to harm.30 Also, there is no consensus on which aspect the intent needs to relate to. The 2018 High Level Group report defines ‘disinformation’ as ‘false, inaccurate, or misleading information designed, presented and promoted to intentionally cause public harm or for profit’, agreeing in so far with the Council of Europe.31 However, it defines ‘misinformation’ as ‘misleading or inaccurate information shared by people who do not recognize it as such’.32 This definition focuses not on the lack of intent to cause harm, but on the lack of knowledge about the falsity of the information. While intent to do harm may provide evidence that the agent in question did not genuinely hold the view expressed,33 the opposite is not necessarily the case. Honesty of belief cannot always be inferred from the lack of intent to do harm. Moreover, deceptive news content that was dis- seminated without malice may also have the capacity to cause harm. Consider the example of the fabricated report about a terror attack in the German city of Mannheim in March 2018. The report described a ‘bloodbath of

29Wardle and Derakhshan, n 7 above, 33. 30W. L. Bennet and S. Livingston, ‘The Disinformation Order: Disruptive Communication and the Decline of Democratic Institutions’ (2018) 33 (2) European Journal of Communication 122, 124; Y. Benkler, R. Faris and H. Roberts, Network Propaganda: Manipulation, Disinformation and Radicalisation in American Politics (OUP 2018), 6.

31European Commission, A Multidimensional Approach to Disinformation, Report of the High Level Group on fake news and disinformation (Publications Office of the European Union, Luxembourg, 2018) <file:/// C:/Users/Irini%20Katsirea/Downloads/Amulti-dimensionalapproachtodisinformation- ReportoftheindependentHighlevelGrouponfakenewsandonlinedisinformation.pdf>

32Ibid. 33Tse Wai Chun Paul v Cheng [2001] EMLR 31, 777.

166 I. KATSIREA

apocalyptic proportions’ in graphic terms. The blog’s editorial team defended its decision to publish the piece, saying it did not intend to spread panic, but to stimulate debate about possible future threats.34 It is questionable whether such a fabricated report could contribute to a debate in the public interest. In any case, false allegations of terror attacks or of other criminal conduct are likely to incite public disquiet and might be punishable.35

Greater definitional clarity would be achieved by focusing on the honesty of belief as to the veracity of the allegations made. That is not to discount the fact that some honestly held, erroneous beliefs may also have the capacity to inflict harm, as in the much-discussed example of anti-vaccination conspiracy theories on social media and their impact on vaccination rates.36 Such instances of misinformation are outside the scope of this article, which defines ‘fake news’ as publications in the guise of news that consistently publish knowingly false content.

As becomes apparent from the discussion so far, ‘fake news’ is not a homo- geneous concept, but can manifest itself in different ways, raising concerns about the future of democracy, but possibly also falling within the remit of civil or criminal law.37 Having attempted to define the term ‘fake news’ more closely, we will now turn to the question whether ‘fake news’ present a threat that needs to be addressed by way of some kind of legal or regulatory action.

Does ‘fake news’ present a threat?

According to Damian Collins, Chair of the Commons Culture, Media and Sport Select Committee, ‘fake news’ could pose a threat to ‘the integrity of democracy’ because large segments of the population who relied on Facebook for their news could be misled, especially at election time.38 This is the more cause for concern as Facebook vies with traditional media as the joint-fifth highest news source in terms of reach, used by 12% of UK adults, while the most-read newspapers, The Sun and the Daily Mail, only have a reach of 6% among UK adults.39 The risk of disinformation, not least by fraudulent

34‘Fake Terror Report Sparks Backlash in Germany’, 26 March 2018 <https://www.dw.com/en/fake-terror- blog-report-sparks-backlash-in-germany/a-43145395>.

35In Germany, the offence of s. 126 (2) StGB penalises the breach of the pubic peace by knowingly pre- tending that the commission of an unlawful act is imminent.

36D. Jolley and K. Douglas, ‘The Effects of Anti-Vaccine Conspiracy Theories on Vaccination Intentions’ (2014) 9(2) PLOS ONE.

37D. O. Klein and J. R. Wueller, ‘Fake News: A Legal Perspective’ (2017) 20(10) Journal of Internet Law 5. 38C. Brinkhurst-Cuff, ‘MPs to Investigate Threat to Democracy from “Fake News”’, 29 January 2017, <www. theguardian.com/media/2017/jan/29/fake-news-mps-investigate-threat-democracy>.

39Ofcom, ‘Written Evidence Submitted by Ofcom to the “Fake News” Inquiry’, March 2017, <https://www. parliament.uk/business/committees/committees-a-z/commons-select/culture-media-and-sport- committee/inquiries/parliament-2015/inquiry2/>.

JOURNAL OF MEDIA LAW 167

articles published on legitimate news sites lookalikes, is considerable.40 This view is shared by the former US President, Barack Obama, as well as by German poli- ticians who aspired to criminalise targeted misinformation aimed at the destabi- lisation of the state.41 They are doubtlessly familiar with Craig Silverman’s finding that, in the final three months of the US presidential campaign, the most popular fake election news stories on Facebook aroused greater public interest than the top stories from major news outlets such as the New York Times or Huffington Post.42 This finding is in line with a Pew Research Center survey according to which 64% of US adults believe that ‘fake news’ cause a great deal of confusion about the basic facts of current issues and events.43

The effect of ‘fake news’ as well as the extent of this phenomenon are con- tentious. Marc Zuckerberg, Facebook CEO, initially claimed that 99% of all of Facebook content was authentic, that ‘fake news’ were the big exception and that it was extremely unlikely that hoaxes changed the outcome of the US elec- tion.44 However, Zuckerberg later changed his mind and vowed to take measures to tackle the problem in an effort no doubt also motivated by the wish to counteract bad publicity.45 Facebook has placed particular emphasis on so called ‘false amplifiers’, which use fake accounts to spread coordinated and/or repeated, rapid posts across multiple surfaces for ideological rather than financial motives, though the two types of motivations can merge. The originators can be automated ‘social bots’ but also human actors. Given the scale of the Facebook platform, even if just 1% of its content consists of fake stories, and if such content only influences a minority of Facebook users, this might still be sufficient to sway election outcomes in marginal seats. Facebook’s current approach to addressing such phenomena focuses on the ‘authenticity of the accounts in question and their behaviours, not on the content of the material created’.46 Still, the question remains whether the moral panic over the threat of ‘fake news’ is an exaggerated one.

40G. Ruddick, ‘Experts Sound Alarm Over News Websites Fake News’ Twins’, 18 August 2017, <https:// www.theguardian.com/technology/2017/aug/18/experts-sound-alarm-over-news-websites-fake-news- twins>.

41‘Barack Obama: Fake News is a Threat to Democracy-Video’, 18 November 2016, <https://www. theguardian.com/media/video/2016/nov/18/barack-obama-fake-news-is-a-threat-to-democracy- video>; ‘CDU Politiker wollen Strafverschärfung’, 13 Dezember 2016, <http://www.spiegel.de/politik/ deutschland/fake-news-unionspolitiker-wollen-strafverschaerfung-bei-gefaelschten-nachrichten-a- 1125611.html>.

42C. Silverman, ‘This Analysis Shows How Viral Fake Election News Stories Outperformed Real News on Facebook’, 16 November 2016, <https://www.buzzfeed.com/craigsilverman/viral-fake-election-news- outperformed-real-news-on-facebook?utm_term=.dg9NjgZYx#.bm29vjXYr>.

43B. Barthell, A. Mitchell and J. Holcomb, ‘Many Americans Believe Fake News is Sowing Confusion’, December 2016, <file:///U:/ManW7/Downloads/PJ_2016.12.15_fake-news_FINAL%20(1).pdf>.

44J. Jackson, ‘Mark Zuckerberg Vows More Action to Tackle Fake News on Facebook’, 13 November 2016, <https://www.theguardian.com/technology/2016/nov/13/mark-zuckerberg-vows-more-action-to- tackle-fake-news-on-facebook>.

45M. Zuckerberg, ‘Building Global Community’, 16 February 2017 <https://www.facebook.com/notes/ mark-zuckerberg/building-global-community/10154544292806634>.

46J. Weedon, W. Nuland and A. Stamos, ‘Information Operations and Facebook’, 27 April 2017, <https:// fbnewsroomus.files.wordpress.com/2017/04/facebook-and-information-operations-v1.pdf>.

168 I. KATSIREA

Allcott and Gentzkow, in their survey about the effects of ‘fake news’ on the US presidential election, found that social media were an important but not the dominant source of political information, and that television remained far more influential.47 Furthermore, they found that less than 20% of respondents remembered ‘fake news’ and less than 10% believed them. This is in line with further research, which suggests that the effect of news on recipients is ordi- narily very limited and short-lived unless exposure is continuous and very one-sided.48 Allcott and Gentzkow also found that undecided voters were less likely to fall for ‘fake news’ than voters with strong political ideologies. Another recent study into ‘fake news’ consumption during the 2016 US pre- sidential campaign also concluded that such consumption was concentrated among a small group of people with the most conservative online information diets.49 Findings from this rapidly growing area of research so far suggest a very limited impact of ‘fake news’ on political choices.50

These findings chime with research on media habits, which attests the existence of so-called confirmation bias, a propensity of individuals to consume news that support their given beliefs. In the online environment, this arguably leads to the creation of so-called filter bubbles created through algorithms on the basis of online behavioural patterns.51 The extent to which these filter bubbles are hermetically sealed is, however, con- tentious. Do social media really lock users in echo-chambers in which they are only exposed to content that matches their own predilections? There is conflicting evidence to the effect that users of social media, aggregators and search engines often enjoy a more diverse and balanced news diet than non-users.52 Recent research into the nexus between internet use and political polarisation shows that such entrenched divisions are most prevalent among demographic groups least likely to use the internet and social media, and that Facebook users are more likely to encounter ideologically diverse news, notwithstanding Facebook being a key avenue of exposure to ‘fake news’

47Allcott and Gentzkow, n 21 above, 223. 48Rasmus Kleis Nielsen, ‘Fake News: An Optimistic Take’, 17 January 2017, <https://rasmuskleisnielsen.net/ 2017/01/17/fake-news-an-optimistic-take/>; S. Hill et al., ‘How Quickly We Forget: The Duration of Per- suasion Effects from Mass Communication’ (2013) 30(4) Political Communication 521–47.

49A. Guess, B. Nyhan and J. Reifler, ‘Selective Exposure to Misinformation: Evidence from the Consumption of Fake News During the 2016 US Presidential Campaign’, <https://www.dartmouth.edu/~nyhan/fake- news-2016.pdf>.

50B. Martens et al., ‘The Digital Transformation of News Media and the Rise of Disinformation and Fake News’, April 2018 <https://ec.europa.eu/jrc/sites/jrcsh/files/jrc111529.pdf>.

51C. R. Sunstein, Republic. Com 2.0 (Princeton University Press 2007). 52N. Newman, ‘Overview and Findings of the 2017 Report, Reuters Institute Digital News Report at <http:// www.digitalnewsreport.org/survey/2017/overview-key-findings-2017/>; W. Dutton et al., ‘Search and Politics: The Uses and Impacts of Search in Britain, France, Germany, Italy, Poland, Spain and the United States’, 1 May 2017 <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2960697>; R. Fletcher and R. K. Nielsen, ‘Is Social Media Use Associated with More or Less Diverse News Use?’, 25 November 2016 <https://rasmuskleisnielsen.net/2016/11/25/is-social-media-use-associated-with-more-or-less- diverse-news-use/>.

JOURNAL OF MEDIA LAW 169

too.53 Doubtlessly, some social media users exhibit narrow, partisan con- sumption patterns. However, it needs to be borne in mind that this might be a reflection of their conscious choices rather than a consequence of perso- nalisation filters imposed on them by social media platforms as implied by the filter bubble theory.

Moreover, the finding of confirmation bias in the case of ‘fake news’ might actually be less of a cause for concern given that only those people would be affected that were already inclined to believe a particular story. Nonetheless, it is not possible to entirely discount the notion that consistent exposure to fake stories can negatively affect political attitudes by intensifying feelings of alienation and cynicism, especially if not moder- ated by a certain degree of consumption of reliable news.54 A recent longi- tudinal study of false news diffusion on Twitter found that falsehood spread farther and faster than the truth, and that it aroused feelings of fear and disgust.55 Particularly during election times, the risk to social cohesion posed by polarisation is not to be taken lightly. Even though it is candi- dates’ politics rather than the media that decide election outcomes,56 the increased circulation of ‘fake news’, can exacerbate societal problems in the long run. And even if ‘fake news’ do not actually influence election out- comes in tight races, it might take up time and resources to rebut these false stories. The risk of misallocation of resources is also prevalent in other contexts such as during terror attacks and natural disasters as well as in financial markets.57 More research needs to be carried out to more authoritatively pin down these risks.

Having come to the preliminary conclusion that ‘fake news’ can potentially pose a threat to societal cohesion and, ultimately, to democracy, the next section will consider whether there are equal or weightier countervailing factors that support the dissemination of news stories that do not map to reality. These factors may arise from the right to freedom of expression under Art. 10 of the European Convention of Human Rights. The question will be posed in the following section as to how far the case law of the Euro- pean Court of Human Rights (ECtHR; Court) supports the propagation of untruthful expression, taking into account, first, the special role the Court

53L. Boxell, M. Gentzkow and J. Shapiro, ‘Is the Internet Causing Political Polarization? Evidence from Demographics’ (National Bureau of Economic Research, Working Paper 23258, March 2017) <https:// www.brown.edu/Research/Shapiro/pdfs/age-polars.pdf>; E. Bakshy, S. Messing and L. Adamic, ‘Exposure to Ideologically Diverse News and Opinion on Facebook’ (2015) 348(6239) Science 1130–32.

54M. Balmas, ‘When Fake News Becomes Real: Combined Exposure to Multiple News Sources and Political Attitudes of Inefficacy, Alienation and Cynicism’ (2014) 41(3) Communication Research 430–54.

55S. Vosoughi, D. Roy and S. Aral, ‘The Spread of True and False News Online’ (2018) 359 Science 1146–51. 56C. Beckett, ‘What Does the Trump Triumph Mean for Journalism, Politics and Social Media?’, 13 Novem- ber 2016 <http://blogs.lse.ac.uk/polis/2016/11/13/what-does-the-trump-triumph-mean-for-journalism- politics-and-social-media/>.

57R. Spearman, ‘Fake News and Financial Market Blues’ (2017) 8 Journal of International Banking and Finan- cial Law 488.

170 I. KATSIREA

accords to the media, secondly, the different levels of protection it grants to political as opposed to commercial speech, and, finally its jurisprudence on untruthful expression.

The case law of the ECtHR and the protection of untruthful expression

The special role of the press

The Court has construed freedom of expression broadly and the exceptions to which it is subject narrowly, recognising in its early case of Handyside v UK that it

‘is applicable not only to “information” or “ideas” that are favourably received or regarded as inoffensive or as a matter of indifference, but also to those that offend, shock or disturb the State or any sector of the population. Such are the demands of that pluralism, tolerance and broadmindedness without which there is no “democratic society”’.58

A central role for the furtherance of democracy is accorded in the Strasbourg jurisprudence to the free press, on which it is incumbent to impart infor- mation and ideas ‘concerning matters that come before the courts’, ‘on politi- cal issues just as on those in other areas of public interest’, on ‘deficiencies in the operation of Government services, including possible illegal activities’.59

The press performs a vital role as a ‘public watchdog’.60

However, freedom of expression under Art. 10 ECHR is not unrestricted and nor are journalists meant to operate in an ethical vacuum. The Court has recognised that journalists who report on matters of public interest are subject to the proviso ‘that they are acting in good faith in order to provide accurate and reliable information in accordance with the ethics of journal- ism’.61 The requirement of accuracy and the obligation of verification are established in the Court’s case law.62

The duty of verification to which journalists are ordinarily subject does not, however, apply to so-called value judgements. The distinction between factual allegations and value judgements is well established in the case law of the Court as well as in many other jurisdictions. It goes back to the case of Lingens v Austria where the Court ruled that ‘The existence of facts can be demonstrated, whereas the truth of value-judgements is not susceptible of

58Handyside v UK (1976) 1 EHRR 737 para 49. 59Sunday Times v UK (1979) 2 EHRR 245 para 41; Lingens v Austria (1986) 8 EHRR 407 para 41; Observer and Guardian v UK (1992) 14 EHRR 153 para 75.

60Observer and Guardian v UK (1992) 14 EHRR 153 para 59. 61McVicar v UK (2002) 35 EHRR 21. 62Tønsberg Blad AS and Marit Haukom v. Norway (2008) 46 EHRR 30; Bladet Tromsø and Stensaas v Norway (2000) 29 EHRR 125; Selistö v Finland (2006) 42 EHRR 8; Fuchsmann v Germany, application no. 71233/13, nyr, para 45.

JOURNAL OF MEDIA LAW 171

proof … As regards value-judgements this requirement [of proof] is imposs- ible of fulfilment and it infringes freedom of opinion itself … ’.63 Nonetheless, even value judgements need to have some factual basis so as not to be con- sidered excessive.64 We will come back to the distinction between facts and value judgements in the following section when we consider the protection of untruthful expression in the national legal orders under examination. For now, it is sufficient to note that the distinction between facts and value judgements seems to point to the idea that there is something particularly harmful to a false statement of fact. You can take or leave an idea, but a false statement makes a claim to authority.

Even though the abovementioned judgements largely concern the written press, the Court has clarified that the same principles also apply to the audio- visual media.65 The question, however, arises whether this special role with all concomitant obligations and privileges attached thereto only applies to the traditional media or also extends to other actors in the new media environ- ment. The Council of Europe Committee of Ministers has adopted a broad definition of a ‘journalist’ as ‘any natural or legal person who is regularly or professionally engaged in the collection and dissemination of information to the public via any means of mass communication’.66 More recently, the Council of Europe developed a number of criteria which new media actors such as bloggers ought to meet if they are to be regarded as ‘media’.67

The Court also favours a broad definition of the ‘media’. In Steel and Morris it held that the obligation to act in good faith in accordance with the ethics of journalism does not only apply to journalists but also to ‘others who engage in public debate’.68 Further, in a case in which it was not clear whether certain internet postings were placed by a professional jour- nalist in his capacity as such in order to provide information to the public or whether they simply expressed his personal opinions as an ordinary citizen in the course of an internet debate, the Court held that the distinction was imma- terial. What mattered was the fact that he disclosed his identity and that the postings were publicly disseminated on a ‘freely accessible popular internet forum, a medium which in modern times has no less powerful an effect than the print media’.69 The obligation to act in accordance with the ethics of journalism applies to journalists and others who engage in public debate

63Lingens v Austria (1986) 8 EHRR 407 para 46. 64GRA Stiftung gegen Rassismus und Antisemitismus v Switzerland, App no 18597/13 (ECtHR, 9 January 2018), para 68; De Haes and Gijsels v Belgium (1998) 25 EHRR 1 at [47]; Prager and Oberschlick v Austria case (1996) 21 EHRR 1 para 37.

65Haldimann and Others v Switzerland [2015] ECHR 215 para 45. 66Recommendation No. R (2000) 7 of the Committee of Ministers to member states on the rights of jour- nalists not to disclose their sources of information, 8 March 2000, Appendix.

67Recommendation CM/Rec (2011) 7 of the Committee of Ministers to member states on a new notion of media, 21 September 2011, Appendix para 41.

68Steel and Morris v United Kingdom (2005) 41 EHRR 22; see also Braun v Poland [2014] ECHR 1419 para 47. 69Fattulayev v Azerbaijan (2011) 52 EHRR 2 para 95.

172 I. KATSIREA

alike, obviously taking the context of the expression into account.70 It does not, however, require them to prove the veracity of their allegations but only to act with due diligence and in good faith.71

The broad definition of the ‘media’ adopted by the Court suggests that new media engaged in the spread of ‘fake news’ would be held to account for failing to act in good faith and to verify untruthful allegations. Traditional media are all the more responsible to check the source and authenticity of pieces of information before picking them up, reproducing and amplifying them. The heightened responsibilities to which traditional media are held to account by way of regulatory or self-regulatory mechanisms support our pre- viously made argument as to the fallacy of their characterisation as ‘fake news’.72

Having discussed the special role of the press and of other actors with a ‘watchdog’ function in the case law of the Court as well as the obligation of verification and its limits, we will now consider the high level of protection afforded by the Court to political as opposed to commercial speech, in an attempt to locate ‘fake news’ within this hierarchy of expression.

‘Fake news’ and the hierarchy of expression

The European Court of Human Rights does not protect all types of speech to the same extent.73 The margin of appreciation that is afforded national auth- orities and the concomitant intensity of review by the Court are influenced by the category of speech involved. A hierarchy of expression is recognised by the Court with political speech at the apex.74 Artistic speech receives less protec- tion, and a restrained review of the balance struck by the national authorities is operated when it is ‘liable to offend intimate personal convictions within the sphere of morals or, especially, religion’.75 Commercial speech receives the lowest level of protection, though still higher than, say, gratuitous insults or hate speech. The Court considers that a wide margin of appreciation is essen- tial in ‘commercial matters and, in particular, in an area as complex and fluctuating as that of unfair competition’.76 This raises the question as to the level of protection in this sliding scale that should be accorded to ‘fake news’, in the narrow sense given to the term in this study. This question is

70See J. Rowbottom, ‘To Rant, Vent and Converse: Protecting Low Level Digital Speech’ (2012) 71(2) Cam- bridge Law Journal 355, 376.

71ibid para 40. 72It is beyond the scope of this paper to discuss the system of UK press regulation, which though fragmen- ted, provides more oversight than currently exists for social media.

73I am grateful for inspiration for this section and fruitful discussion to Lorna Woods who gave a seminar at the Department of Journalism Studies, University of Sheffield on Wednesday 24 May 2017.

74Sunday Times v UK (1979) 2 EHRR 245; Lingens v Austria (1986) 8 EHRR 407; Thorgeir Thorgeirson v Iceland (1992) 14 EHRR 843.

75Wingrove v UK (1997) 24 EHRR 1 para. 58; Otto Preminger v Austria (1995) 19 EHRR 34; Müller v Switzer- land (1998) 13 EHRR 212.

76Markt Intern and Beermann v Germany (1990) 12 EHRR 161.

JOURNAL OF MEDIA LAW 173

pertinent given that an important incentive for the creation of ‘fake news’ stories is the promise of increased attention by readers, which is rewarded by a greater share of the programmatic, algorithm generated advertising pie.77 ‘Fake news’ stories cover a multitude of topics ranging ‘from the harm- less to the deeply dangerous: from dubious and colourful ‘true crime’ tales to stories playing on racial tensions amid Black Lives Matter protests; from fake promises of political concerts to claims of secret political murders – many naming celebrities in their headlines for an extra viral boost’.78 Do such stories qualify as political expression, which deserves the highest degree of protection, or should they rather be categorised as commercial speech in view of the financial motivation involved?

The commercial aim pursued by the publication of ‘fake news’ should not make a difference as such given that profit-making, corporate bodies are unde- niably covered by Art. 10 ECHR.79 As the Court has observed, a difference in treatment depending on the motivation behind the expression in question might fall foul of Art. 14 ECHR. Also, if freedom of expression was restricted to non-profit journalism organisations, this would deprive a large proportion of the press of any protection.80 Nonetheless, the margin of appreciation that is available to member states when regulating ‘fake news’ might be wider if such news were considered to be akin to advertising. To be sure, ‘fake news’ stories do not seek to incite the public to purchase a particular product or service as in the regular commercial context. However, ‘fake news’ publishers create fabricated stories with the predominant aim of selling the promise of increased viewer attention to an ad agency which agrees to contract their inven- tory.81 In other words, ‘fake news’ refer only incidentally to pretend social or political issues, while their main aim is to draw advertisers to their sites.

This situation is arguably comparable to that in the Raëlien Suisse case where the Court held by a narrow majority that the ban of a poster campaign, which intended to attract people to the cause of the Raëlien Movement, an association which believed in the creation of life on Earth by extraterrestrials, was justified. The Court reasoned that the speech in question was closer to commercial speech than to political speech per se as it did not seek to address matters of political debate in Switzerland but had a certain proselytis- ing function.82

77ITN, ‘Written Evidence Submitted by Ofcom to the “Fake News” Inquiry’, March 2017 <https://www. parliament.uk/business/committees/committees-a-z/commons-select/culture-media-and-sport- committee/inquiries/parliament-2015/inquiry2/>.

78J. Ball, Post-Truth: How Bullshit Conquered the World (Biteback Publishing 2017). 79Sunday Times v UK (1979) 2 EHRR 245; Markt Intern and Beermann v Germany (1990) 12 EHRR 161; Autro- nic AG v Switzerland (1990) 12 EHRR 485 para 47; Casado Coca v Spain para 35.

80Markt Intern and Beermann v Germany (1990) 12 EHRR 161 para 25. 81D. Tambini, ‘Fake News: Public Policy Responses’ (London School of Economics and Political Science, Media Policy Project, Media Policy Brief 20), 11 <http://eprints.lse.ac.uk/73015/>.

82Mouvement Raëlien Suisse v Switzerland (2013) 56 EHRR 14 para 62.

174 I. KATSIREA

The Court’s reasoning met with heavy criticism by the dissenting judges who argued that the introduction of a new category of ‘lower-level’, ‘quasi- commercial’ speech diminished the protection of speech without offering compelling reasons.83 The characterisation of the speech in Raëlien Suisse as ‘quasi-commercial’ is problematic, not least in view of the fact that the organisation in question advocated a certain global vision. The Court banned the poster campaign without clearly spelling out the pressing social need that necessitated such a ban. It subscribed to the view of the national authorities that this ban was indispensable to protect health and morals as well as the rights of others, and so as to prevent crime.84 This argument did not, however, seem entirely convincing given that the association in question operated lawfully since 1977. The Court further argued that the ban of the poster campaign was the least restrictive means given that the association could still disseminate its ideas via its website.85 This argument, which is remi- niscent of the reasoning in Animal Defenders International, seems incongru- ous given that it was the reference on the poster to that very website that lay at the root of the problem.

This case may be contrasted to the Hertel case in which the Court found that the Swiss ban on the dissemination of scientific views on the carcinogenic risk of microwave ovens contravened Art. 10 ECHR. The Court argued that the statements in question were not purely commercial given that they touched upon a debate over public health, and hence affected the general interest.86 The Court notably pronounced that

It matters little that his opinion is a minority one and may appear to be devoid of merit since, in a sphere in which it is unlikely that any certainty exists, it would be particularly unreasonable to restrict freedom of expression only to generally accepted ideas.87

The comparison between Hertel and the Raëlien Suisse case suggests that the Court is not prepared to accept all minority opinions to the same extent. Views that are deemed particularly far-fetched, controversial and potentially harmful are more at risk of being marginalised even if the ‘pressing social need’ hurdle is not met.

The question that needs to be answered then is whether ‘fake news’ could possibly be classified as quasi-commercial speech in accordance with this case law. Publishers of ‘fake news’ do not seek to address matters of political or public debate but to skew debate by spreading fabricated stories at the cost of potential confusion and misinformation of the public with intent to

83ibid, joint dissenting opinion of Judges Sajó, Lazarova Trajkovska and Vučinić. 84Mouvement Raëlien Suisse v Switzerland (2013) 56 EHRR 14 para 72. 85ibid, para 75. 86Hertel v Switzerland (1999) 38 EHRR 534 para 47. 87ibid, para 50.

JOURNAL OF MEDIA LAW 175

commercial or political profit. This might suggest that the Court would afford ‘fake news’ a low level of protection, similar or even lower than that afforded publications whose sole purpose is to satisfy the curiosity of their readership concerning the private lives of public figures.88 Further, one could argue that the difference from Hertel lies in the fact that ‘fake news’ do not propagate minority views on controversial topics in the public interest but views that have absolutely no grounding in reality.

Nonetheless, the equation of ‘fake news’ with quasi-commercial speech harbours risks that are not normally present in the case of purely commercial speech. The intricacies surrounding the accurate definition of ‘fake news’ highlight the dangers associated with the labelling of such speech as ‘quasi- commercial’, which could potentially spill over to other types of protected pol- itical expression. These dangers are compounded by the fact that news media are, arguably, not based on absolute, unassailable truth claims, but on socially negotiated processes of truth finding.89 The acceptability of regulation of commercial speech, say misleading advertising, is explicable by the fact that such expression is generally easier to verify.90 The same cannot be said of pol- itical expression. As a result, the characterisation of news as ‘fake’ will be often fraught with uncertainty.

On the other hand, the pernicious effect of ‘fake news’ on the quality of public debate would certainly have to be taken into account when asses- sing the margin of appreciation that should be accorded to states in this area. The line of cases on paid political advertising are relevant in this context. The Court has ruled that even in the case of the most protected political speech, a somewhat wider margin of appreciation needs to be afforded than would normally be allowed in view of the lack of consensus in the area of paid political advertising.91 The effect of such advertising on the quality of political debate has been accepted to be a relevant factor in the case law of the Court. A wider margin of appreciation afforded to the state would not, however, in itself justify the restriction of ‘fake news’ in the absence of a pressing social need. In the following, we will consider the case law of the Court on the criminalisation of untruthful expression so as to reach a conclusion whether the restriction of ‘fake news’ by way of criminal prosecution would likely be deemed compatible with the Convention.

88Von Hannover v Germany (2005) 40 EHRR 1 para 65. 89LSE, ‘Tackling the Information Crisis’, 9. 90M. Hertig Randall, ‘Commercial Speech under the European Convention on Human Rights: Subordinate or Equal?’ (2006) 6(1) European Human Rights Law Review 53, 80; Virginia State Board of Pharmacy v Vir- ginia Citizens Consumer Council 425 US 748 (1976) para 771 n 24.

91TV Vest AS and Rogalandpensjonistparti v Norway (2009) 48 EHRR 51 para 67; Animal Defenders v Inter- national United Kingdom (2013) 57 EHRR 21 para 123.

176 I. KATSIREA

The ECtHR and the criminalisation of untruthful expression

The European Court of Human Rights has passed verdict on the criminalisa- tion of a specific form of lies, namely those falsifying history by way of Holo- caust or genocide denial claims. The Court has two arrows in the quiver that it uses to tackle such forms of revisionist expression: first, the assessment of the legitimacy of interference with freedom of expression under Art. 10 (2) ECHR, which asks whether the restriction is prescribed by law, whether it pursues a legitimate aim, and whether it is necessary in a democratic society; and secondly, the assessment under Art. 17 ECHR, which outlaws any activity aimed at the destruction of the rights and freedoms set forth in the Convention or at their limitation to a greater extent than is provided for in the Convention. The Court has reserved the more radical treatment under Art. 17 ECHR, the so-called ‘abuse clause’, to cases of Holocaust denial on the basis of the characterisation of the Holocaust as a ‘clearly estab- lished historical fact’.92 In its view, the singularity of the Holocaust means that those found guilty of its denial cannot even rely on Art. 10 ECHR. In the Garaudy case, the Court held that the real aim pursued by Holocaust deniers was to rehabilitate the National Socialist regime and to accuse the victims of falsifying history, thus defaming them and inciting to racial hatred of them in a manner that run counter to the very values which the Convention sought to promote.

The Court’s treatment of Holocaust denial cases constitutes a departure from the more moderate stance of the – now defunct – European Commission of Human Rights and is at variance from the Court’s approach to other cases of historical revisionism. The Commission assessed Holocaust denial or tri- vialisation cases on the basis of Art. 10 (2) ECHR, all the while affording member states a broad margin of appreciation and finding the restrictions on freedom of expression justified in order to protect interests such as national security, the prevention of crime or the reputation and rights of Jews.93 In cases, which did not concern the denial or downplaying of the Holocaust, the Court’s stance has been far more protective of freedom of expression. In Lehideux and Isorni the Court distinguished the debate among historians about the crimes of collaboration in France, which forms ‘part of the efforts that every country must make to debate its own history openly and dispassionately’ from ‘clearly established historical facts – such

92Garaudi v France, no. 65831/01, ECHR 2003-IX. 93H., W., P. and K. v Austria, no. 12774/87, Commission decision of 12 October 1989, DR 62, 216; National- demokratische Partei Deutschlands v Germany, Commission decision of 29 November 1995, DR 84, 149. For a detailed discussion of these and further cases see L. Pech, ‘The Law of Holocaust Denial in Europe: Towards a Qualified EU-Wide Criminal Prohibition’ (New York School of Law, Jean Monnet Working Paper 10/09, 2009) <https://jeanmonnetprogram.org/paper/the-law-of-holocaust-denial-in-europe- towards-a-qualified-eu-wide-criminal-prohibition/>.

JOURNAL OF MEDIA LAW 177

as the Holocaust – whose negation or revision would be removed from the protection of Article 10 by Article 17’.94

More recently, in a judgement by the Grand Chamber in the case of Peri- nçek v Switzerland, the Court drew an even more politically sensitive distinc- tion between the Holocaust and other instances of genocide, notably the Armenian Genocide whose historical and legal qualification are controver- sial.95 The majority found that the criminal prosecution of a Turkish man in Switzerland who denied the characterisation as genocide of the atrocities committed against the Armenians in 1915 amounted to a violation of Art. 10. The Court held that the applicant’s comments were political speech made in the public interest and not an incitement to hatred against the Arme- nian community in Switzerland. It refrained from passing verdict on the legal qualification of the events in question as genocide but expressed its unease with the fact ‘that the Swiss courts appear to have censured the applicant for voicing an opinion that diverged from the established ones in Switzerland, and that the interference took the serious form of a criminal conviction’.96

In conclusion, the Court has so far been prepared to accept only one ‘clearly established historical fact’, the Holocaust, and to show zero tolerance as regards its denial or trivialisation. It has shown far greater tolerance as regards debates surrounding other historical events, going as far as to reason that ‘historical research is by definition subject to controversy and dispute and does not really lend itself to definitive conclusions or the assertion of objective and absolute truths’.97 What conclusions can be drawn from this case law as regards the legal treatment of ‘fake news’? Even tough ‘fake news’ rarely concern historic facts but the distortion or fabrication of current events, the Court’s reasoning suggests that it would be on the whole very reluctant to act as the arbiter of truth. The heavy presumption in favour of free speech in the Court’s case law suggests that a criminalisation of ‘fake news’ would hardly withstand the Art. 10 scrutiny.

The discussion of the ECtHR case law shows that member states in the process of crafting a strategy to confront ‘fake news’, need to carefully con- sider the free speech implications of their initiatives. While a wide margin of appreciation might be accorded in the case of fabricated sites created for profit, the question would need to be answered whether there is a pressing social need to curtail their activities. The extent to which untruthful alle- gations could pose a serious threat to public order, for instance by stirring up hatred, or on the contrary contribute to a debate in the public interest,

94Lehideux and Isorni v France [2000] 30 EHRR 665 paras 47, 55; see also Chauvy and Others v France, no. 64915/01, para 69; Monnat v Switzerland, no. 73604/01, para 57.

95Perinçek v Switzerland, no. 27510/08 (ECtHR, 15 October 2015), para 213 et seq; see also Perinçek v Swit- zerland, no. 27510/08 (ECtHR, 17 December 2013), para 114 et seq.

96Perinçek v Switzerland, no. 27510/08 (ECtHR, 15 October 2015), para 280. 97Perinçek v Switzerland, no. 27510/08 (ECtHR, 17 December 2013), para 117.

178 I. KATSIREA

would need to be taken into account. Given that ECtHR judgements are often fact-specific, the outcome will also depend on whether other rights, such as the right to privacy, are engaged. In the following, we will consider the extent to which three jurisdictions – Germany, the UK and the US – protect untruthful expression, and the consistency of the domestic positions of the former two with the ECHR case law. These are not meant to be exhaus- tive accounts but outlines of the basic tenets of the jurisprudential approaches in the three jurisdictions in question.

National jurisdictions and untruthful expression

Germany

The constitutional basis for the protection of freedom of expression in Germany is Art. 5 (1) GG, which provides that:

Everyone shall have the right freely to express and disseminate his opinion by speech, writing, and pictures and freely to inform himself from generally acces- sible sources. Freedom of the press and freedom of reporting by means of broadcasts and films are guaranteed. There shall be no censorship.

At first sight, one would be forgiven for thinking that Art. 5 (1) GG only pro- tects opinions which, in contrast to assertions of fact, are characterised by the subjective attitude of the person expressing himself to the object of the state- ment.98 Actually, Art. 5 (1) GG also protects the expression of facts insofar as they are a prerequisite for the formation of opinions. As a result, the protec- tion of the expression of facts ends where they cease to contribute anything to the formation of opinion. This is deemed to be the case as far as statements of fact are concerned, which are known or proved to be untrue.99 In other words, ‘incorrect information is not an interest that merits protection’.100 This osten- sibly radical position, is in line with ECHR case law in so far as Holocaust denial, the prime example of a demonstrably false statement, is not protected by freedom of expression.101

Nonetheless, the Constitutional Court sounds a word of caution: the duty of truthfulness may not be interpreted in such a far-reaching way that the function of freedom of expression is endangered or harmed out of fear of sanctions.102 Also, the distinction of statements of opinion from assertions of fact is often difficult when the two are commingled so that they can only together form a meaning. The Constitutional Court has ruled that in such a case, the factual and evaluating components should only be divided if the

98BVerfGE 93, 266 (1995) (’Soldiers are Murderers’ case). 99R. Ricker and J. Weberling, Handbuch des Presserechts (6th edn, C. H. Beck, Munich, 2012), 489 para. 5. 100BVerfGE 85, 1 (1991) (‘Critical shareholders’ case). 101BVerfGE 90, 241 (1994) (‘Ausschwitz Lie’ case). 102BVerfGE 54, 208 (‘Böll’ case).

JOURNAL OF MEDIA LAW 179

sense of the statement is not thereby falsified. Where this is not the case, the statement must in the interests of effective protection be regarded as an expression of opinion in its totality and be included within the scope of Art. 5 GG.103 An example is the distinction drawn by the Constitutional Court between Holocaust denial and the denial of German guilt at the out- break of the Second World War. Similarly to the ECtHR’s findings in Lehi- deux and Isorni, the BVerfG considers that statements about guilt and responsibility for historical events involve complex judgements which cannot be reduced to an assertion of facts and hence fall within the scope of freedom of expression.104 This differentiated case law seeks to protect freedom of expression both in the interest of individual personal development and in the interest of the democratic process for which it is of constitutive importance.105 This, in conjunction with the typical opinion/fact conundrum, means that it is unlikely that entire pieces of writing will be excluded from the scope of Art. 5 GG protection as a result of their characterisation as untruthful expression in form of ‘fake news’.106

In Germany, there is no general law proscribing ‘fake news’. Nonetheless, there is an armoury of criminal law norms providing redress from insult (§185 of the German Criminal Code, StGB); malicious gossip (§186 StGB) and defa- mation (§187 StGB). Defence of truth is only possible in the case of the two latter offences (§§186, 187 StGB). The defence of legitimate interests (§193 StGB) can also be put forward as justification for these offences, and can become especially relevant in the context of reporting on suspicions.107 Per- sonality rights also provide civil law protection from ‘fake news’ by way of an analogue application of §§1004, 823 of the German Civil Code, BGB.108

The recent imposition by way of the Network Enforcement Act of new, more drastic obligations upon providers of social media as regards unlawful content breaching specific provisions of the Criminal Code appears to be a shot in the dark. The term ‘fake news’ is not defined in the law and no case of illegal ‘fake news’ had been known to the government prior to the Act’s adoption.109 The untruthfulness of a news report does not necessarily entail its unlawfulness. In any case, the criminal law provisions whose enforcement

103BVerfGE 90, 241 (1994) (‘Ausschwitz Lie’ case). 104ibid; BVerfGE 90, 1 (1994) (‘War Guilt’ case). 105BVerfGE 82, 272 (1990) (‘Coerced Democrat’ case). 106See H. D. Jarass, D. Pieroth, Grundgesetz für die Bundesrepublik Deutschland (12th edn, C. H. Beck, Munich, 2012) Art. 5 para 4a;

107L. Brost, C. Conrad and F. J. Rödder, ‘Einholung und Berücksichtigung der Stellungnahme bei der Ver- dachtsberichterstattung’ (2018) 4 Archiv für Presserecht 287.

108For more detail on these possibilities of legal redress see Holznagel, n. 4 above. 109Act to Improve Enforcement of the Law in Social Networks (Network Enforcement Act) (Gesetz zur Ver- besserung der Rechtsdurchsetzung in sozialen Netzwerken (Netzwerkdurchsetzungsgesetz – NetzDG)) of 1 September 2017, BGBl 2017 Teil I Nr. 61 <www.bmjv.de/DE/Themen/FokusThemen/NetzDG/NetzDG_ node.html>; Friedhelm, ‘Hate-Speech-Gesetz: Regierung kennt keine einzige strafbare Falschnachricht’, 19 April 2017 <www.golem.de/news/hate-speech-gesetz-regierung-kennt-keine-einzige-strafbare- falschnachricht-1704-127370.html>.

180 I. KATSIREA

the law intends to strengthen, are hardly suited to target ‘fake news’.110 The legislator concedes that only a small fraction of complaints, which have not been resolved by the social network operators, and which are sub- sequently directed to the Federal Office of Justice, are justified.111 Nonethe- less, the new law requires that manifestly unlawful content is removed or blocked within 24 hours of receiving the complaint, and that other unlaw- ful content is removed or blocked within a seven-day time limit, with limited exceptions, on pain of fines of up to five million Euros for repeated breaches.112 Even though the law’s rationale – to enforce existing criminal law norms – is convincing, the procedure put in place is likely to lead to over-blocking and to have a chilling effect on freedom of expression.113

German courts have found Facebook to have unjustifiably blocked content in some instances, while not providing adequate redress in others.114 While it is important to fight illegal content on social media, more balanced solutions that adequately protect the rights to freedom of expression and information are needed.

United Kingdom

The United Kingdom has no codified constitution.115 Freedom of expression is protected under the Human Rights Act, which incorporates the individual rights defined in the European Convention on Human Rights into the law of the United Kingdom.116 As in Germany, there is no general law proscribing untruthful expressions in the United Kingdom. However, falsehood and lies are outlawed by way of specific laws that protect individual interests and/or the public interest.117 Lies that harm a person’s reputation are justiciable by way of the law of defamation.118 Further controls are imposed on untruthful expression in the framework of election law, advertising law, consumer

110G. Nolte, ‘Hate-Speech, Fake News, das ‘Netzwerkdurchsetzungsgesetz’ und Vielfaltsicherung durch Suchmaschinen’ (2017) (7) Zeitschrift für Urheber-und Medienrecht 552, 555.

111Draft Network Enforcement Act of 14 June 2017, German Parliament document 18127/27, 4. 112Network Enforcement Act, s. 3 (2), 4 (2). 113Deutscher Bundestag, ‘Ausarbeitung: Entwurf eines Netzwerkdurchsetzungsgesetzes. Vereinbarkeit mit der Meinungsfreiheit’, 12 June 2017 <https://www.bundestag.de/blob/510514/ eefb7cf92dee88ec74ce8e796e9bc25c/wd-10-037-17-pdf-data.pdf>; K.-H. Ladeur and T. Gostomzyk, ‘Das Netzwerkdurchsetungsgesetz und die Logik der Meinungsfreiheit. Ergebnisse eines Gutachtens zur Verfassungsmäßigkeit des Regierungsentwurfs’ (2017) 6 Kommunikation & Recht 390; J. Wimmers and B. Heymann, ‘Zum Referentenentwurf eines Netzwerkdurchsetzungsgesetzes (NetzDG) – eine kri- tische Stellungnahme’ (2017) 2 Archiv für Presserecht 93; T. Feldmann, ‘Zum Referentenentwurf eines NetzDG – eine kritische Betrachtung’ (2017) 5 Kommunikation & Recht 292.

114LG Berlin, 31 O 21/18, 23 March 2018 <https://dejure.org/2018,8228>; LG Hamburg, 324 O 51/18, 30 April 2018 <https://dejure.org/2018,10548>.

115P. Leyland, The Constitution of the United Kingdom. A Contextual Analysis (3rd edn, Hart 2016), Ch. 2. 116Human Rights Act 1998, c. 42, s. 12. 117J. Rowbottom, Written Submission, March 2017 <https://www.parliament.uk/business/committees/ committees-a-z/commons-select/culture-media-and-sport-committee/inquiries/parliament-2015/ inquiry2/>, para. 6.

118Defamation Act 2013 (Commencement) (England and Wales) Order 2013, SI 2013/3027.

JOURNAL OF MEDIA LAW 181

protection law and criminal law.119 The question whether ‘fake news’ is pro- tected by the right to freedom of expression under Art. 10 ECHR and s. 12 HRA can therefore only receive a qualified answer. We will consider two examples from the areas of election law and criminal law.

The decision in the Woolas case suggests that the right to freedom of expression cannot extend to a right to tell lies about a candidate’s personal character in the course of an election campaign.120 The background to the Woolas case was the 2010 general election where the Labour candidate, Phil Woolas, so as to win a hotly contested parliamentary seat in Oldham, made statements as regards another candidate’s personal character or conduct without believing or having reasonable grounds to believe them to be true. He alleged that the said candidate had attempted to seek the electoral support of Muslims who advocated violence, that he had refused to condemn them, and that he had reneged on a promise to live in the constitu- ency. The Divisional Court stressed that the Representation of the People Act 1983 had a narrower scope than Art. 10 ECHR, which allegedly did ‘not extend to a right to be dishonest and tell lies’. It only aimed to punish dishon- est statements in relation to a candidate’s personal character or conduct, not their political character or conduct. As a result, an illegal practice could only be found in respect of the first two statements but not in respect of the last one. Unfounded allegations about the political position of candidates in an election should be left to the electorate to discount. Outlawing them would stifle political debate. The same does not apply to untrue statements about the personal character of a candidate, which can be highly damaging and hard for the public to assess. A remedy under ordinary law would be difficult to obtain in the midst of an election. Their suppression would there- fore be beneficial for the democratic process.121

Is the distinction drawn in s. 106 of the Act between statements about a candidate’s political position and their personal character compatible with the ECHR case law on freedom of expression? The greater protection afforded to the former is in line with the higher rank accorded by the ECtHR to political speech compared to speech whose contribution to a debate in the public interest is questionable, and which might violate personal rights. However, the distinction between personal and political may be on occasion hard to draw.122 A statement about a candidate’s character may also cast doubt upon their suitability for office. The court’s suggestion in

119Representation of the People Act 1983, c. 2, s. 106; UK Code of Non-Broadcast Advertising and Direct & Promotional Marketing (CAP Code), s. 3; The Consumer Protection from Unfair Trading Regulations 2008, Reg. 5; Fraud Act 2006, c. 35; The Awards for Valour (Protection) Bill [HL] 2017–19 that aims to attach criminal sanctions to the unauthorised use of medals is currently going through Parliament.

120R (on the application of Woolas) v Parliamentary Election Court [2010] EWHC 3169 (Admin), para 106. 121R (on the application of Woolas) v Parliamentary Election Court [2010] EWHC 3169 (Admin), para 110. 122J. Rowbottom, ‘Lies, manipulation and elections – Controlling false campaign statements’ (2012) 32(3) Oxford Journal of Legal Studies 507, 528 et seq.

182 I. KATSIREA

Woolas that any inequities should be ironed out at enforcement stage is hardly conducive to legal certainty.

The Privy Council, in a case decided prior to the adoption of the Human Rights Act, also found that space should be preserved for political criticism. It viewed ‘a statutory provision which criminalises statements likely to under- mine public confidence in the conduct of public affairs with the utmost suspi- cion’.123 In 1976, Antigua and Barbuda, independent Commonwealth States, amended their Public Order Act 1972 to the effect that any person who printed or distributed any false statement, which was ‘likely to cause fear or alarm in or to the public, or to disturb the public peace, or to undermine public confidence in the conduct of public affairs’ should be guilty of an offence. The acting editor of the opposition paper was charged with this offence on the ground that he published a supposedly false statement. The paper had reported testimony given to a United States Senate hearing, alleging that the Antiguan government had three years before kidnapped the wife and child of a cocaine trafficker and demanded a ransom for their release.124 The Judicial Committee of the Privy Council reasoned that those who hold office in government must always be open to criticism. It would be a grave impedi- ment to the freedom of the press if journalists could only criticise public officials’ conduct with impunity if they could first verify the accuracy of all facts on which the criticism was based. Moreover, the words ‘likely to under- mine public confidence in the conduct of public affairs’ violated freedom of expression given that they were also applicable to a false statement, which was not likely to affect public order.

In conclusion, whereas false statements, which seek to discredit a political candidate’s character are regarded as wholly intolerable, others which are uttered as a form of government criticism are considered to be an integral part of the press’s watchdog function even if no sufficient verification had been carried out.

In view of the variety of situations in which inaccurate statements are pub- lished and of the open-ended, highly politicised character of the term ‘fake news’, the United Kingdom has rightly been cautious in its endeavour to come to grips with this phenomenon. Differently from Germany, there has been no legislative initiative with the aim of combatting ‘fake news’ in the United Kingdom so far. The House of Commons Digital, Culture, Media and Sport Committee conducted a ‘Fake News Inquiry’, which came to a halt as a result of the May 2017 General Election.125 The Committee

123Leonard Hector Appellant v Attorney-General of Antigua and Barbuda and Others Respondents [1990] 2 WLR 606; [1990] 2 AC 312.

124A. W. Bradley, ‘Case Comment. Press Freedom, Governmental Constraints and the Privy Council’ (1990) Public Law 453.

125Digital, Culture, Media and Sports Committee, ‘Fake News Inquiry’, March 2017 <www.parliament.uk/ business/committees/committees-a-z/commons-select/culture-media-and-sport-committee/inquiries/ parliament-2015/inquiry2/>.

JOURNAL OF MEDIA LAW 183

subsequently resumed its work, focusing on the definition of ‘fake news’; on their impact on the public understanding of the world and on traditional jour- nalism; on the consumption of ‘fake news’ by people of different ages, social backgrounds etc; and last but not least on the all-important nexus between advertising and fake news.126 Interestingly, the Committee’s interim report shifted its emphasis from these concerns to the role and responsibilities of tech companies as well as to issues of data harvesting, particularly in the context of election and referendum campaigns. While these are certainly per- tinent questions, they leave the thorny problem of identifying the individual and collective harms caused by misinformation underexplored. Further evi- dence-based analysis of the views that are considered as ‘pernicious’, and of the harm associated therewith, is needed if freedom of expression is not to be unjustifiably curtailed.

United States

The protection of freedom of expression under the First Amendment is especially far-reaching. A central tenet of the First Amendment is the govern- ment neutrality in the ‘marketplace of ideas’.127 The ‘marketplace of ideas’ metaphor rests on the notion that ‘the best test of truth is the power of the thought to get itself accepted in the competition of the market’.128 The gov- ernment is not meant to discriminate between different viewpoints, and content-based restrictions on speech are viewed with the utmost suspicion.129

Differently from the European jurisdictions examined before, First Amend- ment doctrine refuses to distinguish between true and false ideas within public discourse in line with A. Meiklejohn’s belief in the ‘equality of status in the field of ideas’.130 All ideas are equal because they reflect the equality and autonomy of the citizens holding them.131 Underlying this wide protec- tion of free speech is a deep-seated distrust in the ability of government to draw the right lines between truth and falsity.132 This acute awareness of the dangers of governmental intervention explains the enduring legacy of the marketplace model despite the many criticisms voiced against it over time.133 Despite the weaknesses of this inevitably imperfect metaphor, it

126Digital, Culture, Media and Sports Committee, ‘Fake News’, September 2017 <www.parliament.uk/ business/committees/committees-a-z/commons-select/digital-culture-media-and-sport-committee/ inquiries/parliament-2017/fake-news-17-19/>.

127Hustler Magazine, Inc. v Falwell, 485 U.S. 46, 56 (1988); FCC v. Pacifica Foundation, 438 U. S. 726 (1978). 128Abrams v United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting). 129E. Barendt, Freedom of Speech (2nd edn, OUP 2005), 51. 130A. Meiklejohn, Political Freedom: The Constitutional Powers of the People (Harper 1960), 27. 131R. Post, ‘Participatory Democracy and Free Speech’ (2011) 97(3) Virginia Law Review 477, 484. 132F. Schauer, Free Speech: A Philosophical Enquiry (CUP 1982). 133See S. Ingber, ‘The Marketplace of Ideas: A Legitimizing Myth’ (1984) 1 Duke Law Journal 1; A. Goldman and J. Cox, ‘Speech, Truth and the Free Market for Ideas’ (1996) 2 Legal Theory 1; D. Strauss, ‘Persuasion, Autonomy, and Freedom of Expression’ (1991) 91 Columbia Law Review 334, 349.

184 I. KATSIREA

persists not because truth can necessarily be discovered through the workings of the marketplace, but because the alternative of a state-sponsored truth seems like the less palatable alternative.

In Gertz v Robert Welch, a case that established the standard of liability in private person libel lawsuits, the US Supreme Court accepted that ‘Under the First Amendment, there is no such thing as a false idea’ but went on to recog- nise that there is ‘no constitutional value in false statements of fact’.134 Both intentional lies and careless errors belong to that category of utterances which ‘are no essential part of any exposition of ideas, and are of such slight social value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality’.135 The Court is only prepared to regulate false speech when a cognisable harm is associated with falsehood as in the case of defamation, misleading commercial expression or fraud. As a result, the Supreme Court struck down the convic- tion of a local government official, who falsely claimed to have received the Congressional Medal of Honor, on the basis of the Stolen Valor Act, which criminalised such false statements about military awards, as incompatible with the First Amendment.136 The Supreme Court held that falsity alone without evidence of a specific harm, would not suffice to bring the speech outside the First Amendment.137 The level of scrutiny, strict or intermediate, was contentious in the Alvarez case. However, Alvarez did not concern politi- cal speech. It is well established that ‘content-based restrictions of political speech must withstand strict scrutiny’.138

Notwithstanding the strong protection for political speech under the First Amendment, a state interest in combatting fraud is recognised during election campaigns when ‘false statements, if credited, may have serious adverse con- sequences for the public at large’.139 However, this does not give the regulation of false political speech a carte blanche. In Susan Anthony List v Driehaus the Supreme Court held that an Ohio statute that criminalised certain false state- ments made in the course of a political campaign deterred speech, not least because calculated claims of falsity could achieve maximum disruption of pol- itical opponents by forcing them to divert time and resources in the crucial pre-election period.140 The Supreme Court never fully addressed the statute’s constitutionality, but a federal district court ultimately struck it down.141

134Gertz v Robert Welch, Inc., 418 U.S. 323, 340 (1974); cf. New York Times v Sullivan, 376 U.S. 254 (1964) as regards the higher standard of liability for public persons.

135ibid; Chaplinsky v New Hampshire, 315 U. S. 568, 315 U. S. 572 (1942); see also Garrison v Louisiana, 379 U.S. 64, 75 (1964); Bose Corp. v Consumers Union of US, Inc., 466 U.S. 485, 504, n 22 (1984).

136United States v Alvarez, 132 U.S. 2537 (2012). 137ibid, 2545. 138Boos v Barry, 485 U.S. 312, 321 (1988). 139McIntyre v Ohio Elections Comm’n, 514 U.S. 334, 349 (1995). 140Susan Anthony List v Driehaus, 134 U.S. 2334 (2014). 141List v Ohio Elections Com’n, 45 F. Supp. 3d 765 (2014).

JOURNAL OF MEDIA LAW 185

Similarly, the United States Court of Appeals for the Eighth Circuit struck down a statute that proscribed knowingly false speech about ballot initiatives given that it tended to perpetuate the very fraud it was allegedly designed to prohibit, and risked chilling protected political speech without adequately explaining why counter-speech was not an equally, if not more effective remedy.142 At the time of the Driehaus decision, sixteen US states had laws prohibiting false campaign speech. It remains to be seen whether these laws as well as others in the pipeline will survive constitutional scrutiny.143

In conclusion, untruthful expression in the form of ‘fake news’ would only fall foul of the First Amendment protection if there was proof of some legally cognisable harm. The abridgement of free speech would hence be justified in the case of the incitement of imminent lawless action, as in Justice Oliver Wendell Holmes’s famous example of ‘falsely shouting fire in a theatre and causing a panic’, or in the case of campaign lies about individual candidates that are accompanied by the mens rea necessary for defamation.144 The wide- spread consensus, in line with traditional First Amendment doctrine, is that untruthful expression should not otherwise be regulated, but faith should be placed on more truthful expression to counter the falsehood and prevail in the marketplace of ideas.145 In the words of Justice Brandeis, ‘if there be time to expose through discussion the falsehood and fallacies, to avert the evil by the processes of education, the remedy to be applied is more speech, not enforced silence’.146

Conclusion

This article has analysed the nexus between the right to freedom of expression and the ever prevalent spread of untruthful content in our complex infor- mation ecosystem. ‘Fake news’, narrowly defined here as ‘the knowing and consistent publication of predominantly false information in the guise of news with intent to deceive’, holds potential risks for the formation of individ- ual and public opinion, for societal cohesion and for the future of democracy. Discussions rage about how to tackle the new/old phenomenon of ‘fake news’ and a myriad of initiatives spring up aimed at better spotting, highlighting or down-ranking such content, at eliminating financial incentives for its creation or at otherwise suppressing it, or on a more positive note, at strengthening

142281 Care Committee v Arneson, 766 F.3d 774 (8th Cir. 2014). 143J. Sellers, ‘Legislating Against Lying in Campaigns and Elections’ (2018) 71(1) Oklahoma Law Review 141, 143; see Deceptive Practices and Voter Intimidation Prevention Act of 2018, 26 July 2018 <https://www. congress.gov/bill/115th-congress/senate-bill/3279/text?format=txt>.

144Schenck v United States 249 US 47 (1919); see A. Sarat, Law and Lies. Deception and Truth-Telling in the American Legal System (CUP 2015), 178.

145P. Callan, ‘Sue Over Fake News? Not So Fast’ <https://edition.cnn.com/2016/12/05/opinions/suing-fake- news-not-so-fast-callan/index.htm>; contra A. Hundley, ‘Fake News and the First Amendment: How False Political Speech Kills the Marketplace of Ideas’ (2017) 92 Tulane Law Review 497.

146Whitney v California, 247 US 352 (1927).

186 I. KATSIREA

trustworthy news outlets and at improving media literacy. What some of these initiatives pay insufficient regard to is the right to free speech that is at stake. Our discussion of the ECHR case law shows that ‘fake news’ outlets’ blatant disregard of the truth would fall foul even of the most basic ‘due diligence’ and ‘good faith’ obligations that go hand in hand with engagement in public debate. The Court might even afford member states seeking to regulate such outlets a wider margin of appreciation, comparable to that applicable in cases of problematic commercial expression and of paid political advertising.

However, this alone would not justify the restriction of ‘fake news’ unless if there was a pressing social need. Whereas the introduction of new criminal sanctions on untruthful expression is problematic, there is ardent need for more research to be carried out so as to enable us to more accurately identify risks, to evaluate the argument for regulatory intervention, and to adopt responses that are proportionate to these risks.147 This finding is supported by our examination of case law from Germany, the UK and the US. Despite their diverse constitutional traditions, these jurisdictions share a pre- paredness to curb untruthful expression only in specific settings when it does not contribute to debate in the public interest and when harm to private or public interests can be made out, a yardstick for this being the illegality of the content in question. This is also reflected in the new German Network Enforcement Law, which only fines platforms for hosting unlawful content. Existing laws already provide a powerful armoury to fight disinformation in form of inter alia defamation or election fraud. Innovative solutions are needed for smart enforcement in the online environment that is respectful of users’ fundamental rights. Outsourcing of law enforcement to private plat- forms is problematic if effective oversight mechanisms are not in place. Also, such platforms need to adopt a more transparent modus operandi as regards their content regulation and dissemination processes, and to strengthen their self-regulatory strategies.148 The steps taken by Facebook and Google to tweak their advertising business model so as to discourage the financial exploitation of ‘fake news’ sites go in the right direction.

A more wide-reaching effort to clamp down on ‘fake news’ might be defensible on moral grounds in the sense that deliberate falsehoods in communication are reprehensible.149 One might even be tempted to argue that such utterances are undeserving of the main justifications of free speech, the arguments from truth and from citizen participation in a

147Such an evidence-based approach underlies the recommendations made in the 2018 High Level Group report, n 31 above; see also The United Nations (UN) Special Rapporteur on Freedom of Opinion and Expression et al., ‘Joint Declaration on Freedom of Expression and “Fake News”, Disinformation and Pro- paganda’, 3 March 2017 <https://www.osce.org/fom/302796?download=true>.

148V. Marda and S. Milan, ‘Wisdom of the Crowd: Multistakeholder Perspectives on the Fake News Debate’, 21 May 2018 <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3184458>.

149T. Gibbons, ‘“Fair Play to All Sides of the Truth”: Controlling Media Distortions’ (2009) 62(1) Current Legal Problems 286, 299.

JOURNAL OF MEDIA LAW 187

democracy.150 The dissemination of ‘fake news’ does not aid the discovery of truth and might challenge democratic governance by disorienting citizens and distorting election outcomes. Even though the improvement of online news quality is a desideratum, the adoption of new criminal law sanctions to curb the flow of ‘fake news’ should be resisted. Not only would it ignore Milton’s advice: ‘Let her and falsehood grapple; who ever knew Truth put to the worse, in a free and open encounter’ and Mill’s forewarning: ‘however true it may be, if it not fully, frequently and fearlessly discussed, it will be held as a dead dogma, not a living truth’.151 More worryingly, it would mean entrusting governments or technology corporations without edi- torial culture with drawing the lines between truth and falsity, thus jeopardis- ing the integrity of democracy with the very means that are meant to uphold it. Restraint with the suppression of ‘fake news’ is advisable not because of such news’ inherent value but because of the ancillary consequences of its restriction for protected truthful expression.

Acknowledgments

I am grateful to Jackie Harrison, Jacob Rowbottom and Kyu Ho Youm for their insightful comments and encouragement. The usual disclaimer applies. All websites in this article were last accessed on 12 October 2018.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributors

Irini Katsirea is Reader in International Media Law at the University of Sheffield, UK. Her current research focuses on the challenges posed to press regulation by digitalisa- tion and convergence and on the limits of free speech online. She is principal inves- tigator of a collaborative research project for IPSO on ‘Examining the impact of IPSO on editorial standards and complaints handling’ (2017–18).

150Barendt, n 105, 7, 18. 151J. S. Mill, On Liberty (Batoche Books, Kitchener, 2001), 34; J. Milton, Areopagitica; and, of Education (Clar- endon Press, Oxford, 1973), 38; contra S. Schiffrin, Speech Matters. On Lying, Morality and the Law (Prin- ceton University Press, Princeton, 2014), 140 et seq.

188 I. KATSIREA

Copyright of Journal of Media Law is the property of © Hart Publishing, Oxford and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.