When Donald Trump suggested he would build a wall on the Mexican border, Hillary Clinton retorted, “How high does a wall need to be to keep out the internet?” Today, radicalisation takes place in bedrooms, in libraries, on mobile phones. Connectivity and globalisation cannot be stopped – nor should they. But how can we stop the oncoming traffic of internet radicalisation?
The emergence of ISIS and its use of the internet for recruitment and propaganda has been a stark reminder of how the web can be a platform for dangerous ideas. Time and again, we have seen how radical thinking online has violently manifested itself in the offline world.
To grasp the scale of this challenge, and to be adequately placed to combat it, we need to diagnose the extent of Islamist extremist material on the internet.
Research into online radicalisation has mainly focused on two areas of enquiry: the role of social media and ISIS’ activity online. But there is much more to the internet than social media. This study will endeavour to shed light on how accessible extremist content is outside of social media, with a particular focus on the role played by the search engine Google. Initiatives for better understanding extremism on the internet have predominantly been led by experts in extremist ideology or the sociological aspects of radicalisation. Technology firms, key stakeholders in this fight, have played a less prominent role.
This study set out to look beyond the usual suspects of Facebook, Twitter, and Telegram. We wanted to gain a greater understanding of the broader reach of extremist ideology online. This focus on social media and bots, the deep and dark web, and how ISIS is harnessing these, has dominated the conversation. Not enough attention has been paid to the wider online landscape, where a broad array of extremist material has flourished, unchallenged.
In a similar vein, recent attention on online extremism has focused on violent material and a handful of groups. But there is an entire spectrum of content online, which expresses ideas that undermine the rule of law and spreads intolerant attitudes. This includes non-violent Islamist groups who call for the downfall of democratic systems and the introduction of a caliphate. It also includes Salafi websites that insist on death for apostates.
ISIS may seem like the apex of Islamist extremism, but it shares an ideology with a number of groups and individuals who seek the same objectives and share a worldview. The researchers set out not only to identify jihadi content online, but also the extent to which nonjihadi extremist content is accessible to users online. Specifically, we wanted to find out how easily the average user could access extremist material.
The debate around how best to combat extremist content online almost inevitably raises questions about the threat posed to civil liberties. While the removal of child pornography or graphic violence is widely agreed upon by governments and technology companies, efforts to counter online extremism should not be used to curb individual rights. Counter efforts must adopt a proactive approach and seek to drown out extremist content, rather than focusing solely on removing it.
This research centred on three layers of analysis: Firstly, understanding the keywords, or search terms, people use to find information on Google; secondly, looking at the data demonstrating links going into a selection of known extremist websites in order to understand their relationships with other websites; and thirdly, analysing the content of Google search results pages to understand the placement of extremist and counter-narrative content within search results for relevant keywords. The areas of analysis represent important aspects related to the broader internet and, when combined, gave the opportunity to get a snapshot of extremist content beyond the realm of social media.
The results produced by this multi-faceted approach provide an overview of extremist content online and shed much-needed light on the impact and effectiveness of online counter-narrative efforts.
As part of the study, we looked at:
The average monthly number of global searches conducted in Google (search frequencies) for 287 extremism-related keywords, 143 in English and 144 in Arabic.
Regional search frequencies in 33 regions, including six US cities, eight UK cities, and 11 countries from the Middle East and North Africa.
The first two search engine results pages for 47 keywords to determine rankings of extremist and counter-narrative content, looking at a total of 870 web pages.
The linking data of 45 websites containing extremist content, in order to understand inter-website relationships and search engine optimisation (SEO) efforts.
A wide range of extremist content is available online. This study found a broad array of extremist content on websites, including violent and non-violent publications. Extremist views on sectarianism, apostasy, and conspiratorial attitudes towards the West – ideas that permeate much of ISIS' output – feature on mainstream Islamic websites. We found that, of the extremist content accessible through these specific keyword searches, 44 per cent was violent, 36 per cent was non-violent, and 20 per cent was political Islamist content, i.e. non-violent content propagated by, or in support of, a known Islamist group with political ambitions.
Web searches are a gateway to violent extremist content. The average, interested internet user requires nothing more than a simple Google search to gain access to extremist publications from groups like ISIS and al-Qaeda. Whether through analysis sites or otherwise, jihadi content is accessible via Google, without the need for social media. From our sample, we found there are on average in the region of more than 484,000 Google searches globally, and at least 54,000 searches in the UK alone, each month for keywords that return results dominated by extremist material. While a wide range of people may have conducted some of these searches, including journalists, researchers, and students, the risk posed by the prevalence of extremist content in these search results is of concern.
High-risk keywords go unchallenged by counter-narratives. Despite the emergence of online counternarrative initiatives in recent years, these narratives are desperately weak in their presence in search engine results pages (SERPs). They do not sufficiently contest or dominate extremist ideas online. After analysing the SERPs for 47 relevant keywords, we found that counter-narrative content outperformed extremist material in only 11 per cent of the results generated. The SERP analysis in this study showed that the content returned for popular keywords often associated with extremism online, such as ‘ISIS,’ ‘Islamic State,’ and ‘jihad’ in fact pose no threat at all. However, extremist content that appears in searches for words that have appeared in discourse in recent years such as ‘caliphate’ and ‘Dabiq’ – the name of ISIS’ English-language magazine – goes unchallenged. Efforts to counter extremism online are lagging behind in agility and diversity.
Mainstream Islamic websites are hosting extremist content. Through the analysis of SERPs and website links it became evident that a number of sites purporting to present legitimate Islamic scholarship also host extremist material. This is true of forums, ‘Q&A’ websites, and online repositories that contain books and lectures by known jihadis such as Anwar al-Awlaki, a US-born jihadi ideologue for al-Qaeda in the Arabian Peninsula (AQAP), and Abdullah Azzam, an early al-Qaeda ideologue and chief controller of the Arab-Afghan Mujahedeen movement. This material is available without any warning or safeguards in place. In some cases, when users were directed to sites for Islamic content, such as translations of the Quran or books of Hadith literature, they were in fact entering sites hosting politicised Islamist and jihadi content. Users who trust a site’s legitimate Islamic content may become vulnerable to extremist ideas they encounter on that same site.
Non-violent Islamists have a strong online presence. Hizb ut-Tahrir, a global pan-Islamic movement that seeks to establish an Islamic state, has a very strong online presence and dominates the results for a number of the keyword searches in our sample. The group operates a series of websites, including its central site and regional affiliate websites that support Hizb ut-Tahrir’s global aims. Furthermore, our research showed that political Islamist content by Hizb ut-Tahrir, its affiliates, and its supporters accounted for 20 per cent of all the extremist content identified.
Counter-narratives are lagging, but Muslim efforts dominate. Counter-narrative efforts are not challenging the extremist content found in seach engine results pages, with efforts appearing in only 43 of the 870 results analysed, just five per cent of the total. However, of the counter-narratives identified, 91 per cent were Muslim-led. This highlights the efforts taken by Muslims to address the rising tide of extremist ideology online. It is estimated that close to three billion people have access to the internet around the globe, a number expected to swell to more than seven billion by 2020. The need to address the issue and safeguard internet users has never been more pressing.
This report provides a brief, albeit targeted, reflection of a tiny slice of the constantly growing and evolving world wide web, and access to extremist content on it. More importantly, our research is designed to inform governments, technology companies, and civil society groups fighting online extremism about how accessible and prolific this material is online.
This ever-evolving landscape, accessed by almost half of the world’s population, must be secured in order to prevent extremism online from manifesting offline. We need a cohesive, integrated, and concerted effort to root out extremist ideologies on the internet. But we will only succeed if all those involved fully understand the scale and implications of this unprecedented problem.