Click here to read our latest report: Going Dark: The Inverse Relationship between Online and On-the-Ground Pre-offence Behaviours in Targeted Attackers

The Internet and Radicalisation Pathways: Technological Advances, Mental Health and the Role of Attackers

The Internet and Radicalisation Pathways: Technological Advances, Mental Health and the Role of Attackers
16th January 2023 Dr. Jonathan Kenyon
In Insights


In recent years, one of the primary concerns for policymakers and practitioners working in the field of counterterrorism has been over the role of the Internet in the process of radicalisation. There is little doubt that the growth of the Internet, with its ability to connect people and facilitate the dissemination of ideas and information, has had a significant impact on the accessibility and flow of extremist views. It is clear the online domain provides a platform for those with extremist views to communicate, collaborate and convince others to engage with extremism. However, even if the Internet does play a role in the development of extremist views, how often does this translate into acts of extremist offending that result in arrest and conviction? 

We recently published a report as part of the UK Ministry of Justice Analytical Series investigating the role of the Internet in the radicalisation pathways of convicted extremist offenders in England and Wales.  The study was undertaken by the author, working for His Majesty’s Prison and Probation Service (HMPPS), alongside Dr Jens Binder from Nottingham Trent University and Dr Christopher Baker-Beall from Bournemouth University. The study was an extension of work previously reported by Kenyon, Binder and Baker-Beall (2021), which investigated the same topic.

Our new study was informed by a data set of 490 convicted extremists, representing close to the entire convicted extremist population in England and Wales, from October 2010 until December 2021.  The analysis focused on 437 cases who were identified as ‘Radicalised Extremists’. By this, we mean those found to hold extremist views prior to committing their offence and where there was evidence they had engaged in extremist activity in the outside world.

The study was unique in the sense that whilst other studies rely on open-source data, our study was able to access closed-source data in the form of Structured Risk Guidance (SRG) and Extremism Risk Guidance (ERG22+) assessment reports for each offender. These risk reports were completed by either Registered Psychologists or qualified Probation Officers who had undertaken the same national training in the use of the assessment. Not only did assessors have access to restricted information sources such as Police and Court reports, but in the majority of cases, they also had the opportunity to directly interview the subject of the reports.  The focus of the analysis was on a comparison between three different radicalisation pathway groups: those who primarily radicalised online; those who primarily radicalised offline; and those subject to radicalising influences in both domains – known as the hybrid group.   

For this GNET Insight, our report findings are grouped under four headings: 1. Role of the Internet and technological advances; 2. Pathway group differences; 3. Relevance of mental illness, neurodivergence and personality disorder; and 4. Pathway groups and risk, including the role of attackers.   

Role of the Internet and Technological Advances

One of the most important findings from our work is that radicalisation through online influences has been steadily increasing, while the opposite development holds true for offline influences.  We find a shift has occurred in recent years (2019-21) from a hybrid or mixed pathway being most dominant for convicted extremists in England and Wales, to radicalisation taking place primarily online. This finding contrasts with conclusions drawn from other studies, where a distinction between online and offline radicalisation has been described as afalse dichotomy and radicalisation is reported as taking place across both domains for most individuals. 

This shift may in part be explained by restrictions implemented because of the COVID-19 pandemic in early 2020. As a result, we recognise it is possible the hybrid pathway may potentially return to being the most prominent route to radicalisation in the future. When examining sub-groups, we found that the increased prominence of the Internet in radicalisation pathways was most marked for convicted women and those over 25. Across ideologies, the Internet was found to play a more prominent role in pathways for Islamist extremists, those affiliated with the Extreme Right Wing and those within the Other Political group, which comprised largely of individuals described as anti-establishment or supporting a far-left ideology, or those affiliated with nationalist or separatist movements. Animal rights activists were the exception, with in-person contact remaining a key feature within their radicalisation pathways over time. 

As we found in our earlier study, the types of websites, platforms and applications used by convicted extremists had changed over time. Although a steady decline was noted in the use of extremist websites and standard communication applications, an increase was found in the use of online forums, open social media platforms and encrypted applications. Somewhat surprising, in contrast to findings in the literature, was the low frequency of reported use of the dark web, whilst reported new developments relating to online extremist activity such as online gaming and use of imageboards were overall rather infrequent. The heavily publicised act of livestreaming terrorist attacks, perhaps most infamously employed during the Christchurch mosque shooting in 2019, was not reported as relevant for any cases in the sample.

Pathway Group Differences – Profiles, Offending History, Socialisation and Online Activities

When comparing pathway groups, differences were found in terms of demographic profiles, offending history and degree of socialisation of offenders. Those who primarily radicalised online were characterised as being most likely to commit a solely online, non-violent extremist offence. They were also least likely to be socially connected to other extremists offline.  In contrast, those who primarily radicalised offline were typically older, with a convicted and often violent offending history.  Offenders within this pathway group were also most likely to have committed a violent extremist offence resulting in their most recent conviction, be part of a group of four or more members, and least likely to follow an Islamist extremist ideology. Those belonging to the hybrid group were most likely to be part of a small cell, consisting of two or three people, and were most likely to have committed previous extremist offences. 

Differences were also found across the three pathway groups in terms of the types of extremist online activities undertaken. Importantly, those who primarily radicalised online were more likely to learn from online sources, interact with like-minded others online and use open social media platforms within the context of their offending, compared to those who primarily radicalised offline. Differences were also found when compared with the hybrid group, with those who primarily radicalised online more likely to use open social media platforms and generate their own extremist propaganda.

Relevance of Mental Illness, Neurodivergence and Personality Disorder

The role of mental illness, neurodivergence and personality disorder were investigated under the broad category of ‘mental health issues’ to reflect how these varied conditions are captured within ERG22+ assessments. This area received attention in response to a reported knowledge gap around the mental health needs of convicted extremists and against a background of media reports quoting the “staggeringly high” proportion of people with autism being referred to the UK Prevent programme. We found that a third of the sample was reported as having either a mental illness, neurodivergence or personality disorder/difficulties. Estimates for the general population suggest that 1 in 6 adults in England have a common mental disorder, whilst surveys have found that 1 in 8 people aged 16 or over screened positive for any type of personality disorder. Whilst up-to-date data for the prison population are hard to come by, estimates are generally higher. Within our sample, the most common disorders were autism spectrum condition (ASC), depression and personality disorder. Importantly, all three were most prevalent in those who primarily radicalised online. 

Pathway Groups and Risk, including Role of Attackers

The different types of methods employed by extremist offenders suggest there are marked differences in terms of the risk presented by individuals depending upon the pathway taken. For example, differences were found in professional assessments contained within ERG22+ reports, pertaining specifically to an individual’s overall level of engagement with an extremist group or cause at the time of offending, and their intent and capability to commit violent extremist acts. Those who primarily radicalised online were considered the least identified with or committed to an extremist group or cause, and least willing and able to perpetrate acts with the potential to cause serious harm. Coupled with the finding that the online pathway was most prominent for those sentenced between 2019-21, and indications this pathway group are most likely to have committed a solely online, non-violent offence, this may suggest the overall threat of serious harm from terrorist offending in England and Wales is starting to diminish.

Those exposed to radicalising influences both online and offline were assessed by professionals as being the most engaged at the time of offending and presenting with the highest level of intent to commit violent acts in support of extremist causes. To further evidence the impact of pathways when considering risk, those in the hybrid group were found to be most likely to have been convicted for further terrorism offences when reviewing future proven offending outcomes. This poses several interesting questions in relation to whether this group may be more resistant to rehabilitation efforts or present additional challenges for professionals. For example, it may be the case that risk management strategies currently available for extremist offenders, following release into the community, are not as robust for those with histories of engaging in extremist activity in both the online and offline domains.

When focusing on the sub-set of the sample identified as attackers (n = 137) due to the high levels of publicity they receive and the concern they cause to the public, online planned action behaviours were found to differ across pathway groups. Compared with those who primarily radicalised offline, both pathway groups exposed to online influences were more likely to engage in attack preparation online. In terms of specific behaviours, those in the hybrid group were more likely to identify targets online, whilst those who primarily radicalised online were more likely to use the Internet to signal attacking intent and in doing so, disregard operational security. 

What was particularly noteworthy from a risk management perspective was that those attackers who primarily radicalised online were least likely to have plots that progressed beyond the planning stage and most likely to have been foiled. This finding was consistent with those of other studies, which focused on cases across several Western countries. We suggest this provides some encouragement to the Police and Security Services that potential terrorists often leave online footprints that can be found, aiding the apprehension of potential terrorists and supporting ongoing investigations. Our study also found that attackers who primarily radicalised offline were most likely to carry out a successful attack and least likely to have been foiled, therefore proving the hardest to detect.   


What do these findings mean for the future of counterterrorism policy and practice?  It is clear from our study that the role of the Internet in radicalisation to extremist offending will continue to be an important issue moving forward.  To counter this threat, we suggest that online responses should be front and centre of counterterrorism efforts in England and Wales, given the increase in prominence of the Internet in radicalisation pathways. However, any measures employed must reflect the heterogeneity of the audience. The changes in Internet use over time, coupled with the rapidly changing online technological landscape, suggest that research efforts should focus on how emerging platforms and applications are being used to inform response strategies. Given the variation of platforms and applications being utilised by extremists, in what has been described as an online ecosystem, multi-platform and multi-stakeholder responses are likely to prove most effective. One example is the Global Internet Forum to Counter Terrorism (GIFCT), which utilises a shared ‘hash database’ and relies upon coordination by member platforms to support content moderation.

Having found a third of the convicted sample had either mental illness, neurodiversity or personality disorder/difficulties, with this particularly prominent for those radicalised online, specialist support is likely to be beneficial for this cohort, particularly for ASC, depression and personality disorder. As those who primarily radicalise offline appear most likely to carry out a successful attack and prove hardest to detect, we suggest the Police and Security Services should pay particular attention to the offline activities of those assessed as a potential threat, with reasonable concerns of potential attack planning being acted upon at an early stage. For those who primarily radicalise online, diversionary solutions at the point of sentencing or even arrest may, in some instances, prove to be a more effective response than automatic prison sentences. This may be most applicable to those who are considered vulnerable, have engaged with an extremist group or cause solely online, and where their involvement is seen as no more than peripheral.

Dr Jonathan Kenyon is a BPS Chartered and HCPC Registered Psychologist. He currently works as a National Specialist Lead within the HMPPS Counter Terrorism Assessment and Rehabilitation Centre (CT-ARC). His research interests include pathways to lone-actor terrorism and exploring the role of the Internet in radicalisation and extremist offending. Recent publications include Lone-actor terrorism – A systematic literature review (Kenyon, Baker-Beall & Binder, 2021), The Role of the Internet in Radicalisation and Offending of Convicted Extremists in England and Wales (Kenyon, Binder & Baker-Beall, 2021) and Terrorism and the Internet: How dangerous is online radicalization? (Binder & Kenyon, 2022).

Twitter handle: @jon_kenyon1