Most convicted terrorists in Britain were turned to extremism by the internet, with half of those radicalised online having some problems with mental health, personality disorders, depression, or autism, the most authoritative study of its kind has found.
The study for the Ministry of Justice, released on Thursday, examined official risk assessments of every convicted terrorist in prison since 2010 in England and Wales. The majority were radicalised at least in part online – a trend caused by technology and the tactics of groups such as Islamic State.
While those incited to carry out or support violence online make up the majority of terrorists, their plots, according to the study, are less likely to succeed.
The research challenges conventional wisdom that the growth of internet radicalisation by Islamists and the extreme right allows terrorism to fester out of sight of the security services and police.
The researchers from the prison and probation service, as well as Nottingham Trent and Bournemouth Universities, were given access to Extremism Risk Guidance 22+ forms. These are written for the prison and probation services and assess the commitment to terrorism, as well as the danger posed by convicted terrorists. The forms include material from counter-terrorism policing.
For all convicted terrorists in jail from 2010 to 2021, 490 records were retrieved and 437 studied in detail. Nine out of 10 terrorist prisoners were male, 44% were aged 25 or under when convicted, and seven out of 10 were British-born. The study finds 4% were convicted for animal rights violent extremism, 18% for extreme rightwing terrorist offences, and 72% for Islamist-incited offences.
The study reveals the growing importance of the internet to its current-day dominance in radicalisation. In 2013–15, 43% of those in prison for terrorism were radicalised wholly or in part online. That figure rose to 84% in 2016–18, and between 2019–21 was 92% of those convicted. The latter figure may have been further inflated by Covid lockdowns.
From 2015 onward, IS launched an intense online propaganda war, with the far right mimicking its strategy as they tried to gain recruits.
For those radicalised online between 2010 and 2021, 28% had previous convictions for any crime, and 15% of these were for violent offences. None had a past terrorism conviction, 4% had a past conviction for a terrorism-related offence, 42% has a “strong presence” of mental health issues, neurodivergence and personality disorder, and 9% had these partly present.
In contrast, it is estimated one in six adults in England have a common mental disorder, and one in eight people aged 16 or over screened positive for any type of personality disorder.
The study looked at those convicted of terrorism, and its findings are echoed by other work on mental health and terrorism. The emerging importance of mental health as a factor has also been seen by the Prevent programme, which tries to stop people from being radicalised in the first place.
Last year the Guardian revealed that up to seven in 10 people referred to the scheme may experience mental ill health or other vulnerabilities that could leave them susceptible to falling for propaganda from violent extremists. Those involved in Prevent believe such psychological problems are much more of a potential factor than first thought.
Terrorists whose records were examined for the study include Roshonara Choudhry, convicted in 2010 of trying to assassinate the Labour MP Stephen Timms. The student wanted to die as a martyr after watching more than 100 hours of extremist video sermons on YouTube. It also studied the ERG of Darren Osborne, convicted of the 2017 attack on worshippers leaving Finsbury Park mosque, who self-radicalised in three weeks.
The study showed that, for those radicalised on the internet, 84% were convicted for non-attack offences such as spreading propaganda or fundraising; 16% for offences plotting violence; 85% were lone actors; 7% had suffered a head or brain injury; 75% were convicted of offences supporting the Islamist cause and 25% other ideologies, the biggest of which was fuelled by extreme rightwing beliefs.
For those radicalised primarily face to face, half had past convictions, more than a third for violence and 5% had past terrorism convictions; 19% had any sort of mental health issue or vulnerability, and 51% of these offenders were convicted for planning attacks. Only 6% were lone actors, 58% were Islamist, and 42% had other ideologies.
The report says: “For those who primarily radicalised online, the most common types of plots included the use of an improvised explosive device (IED, 65%), a bladed weapon (24%) or a vehicle (12%).
“Only the minority of plots moved from planning to the execution stage (29%), with 18% of plots successfully completed. For this pathway group, all thwarted plots were disrupted by the police or security services (100%), suggesting that the online traces of those who primarily radicalised online make it more difficult for them to progress substantially in their attacks and bring them to the attention of the police or security services more readily.
“This interpretation is also supported by the finding that attackers who primarily radicalised online were most likely to signal their attacking intent compared to the other pathway groups. These findings also counter the popular notion that the Internet helps create an undetectable threat of lone actors.”
Those radicalised online were less committed to the ideological cause, and assessed as less capable than those recruited face to face.
The three broad types of radicalisation are online, real world and a mix of both. The report finds recruits to extreme rightwing ideologies, using online forums including Iron March and Fascist Forge. An app originally designed for gamers called Discord was also sometimes used. The report notes that some incited to support Islamist terrorism were radicalised by games such as Call of Duty.
The MoJ said the views expressed in the report are those of the authors and “are not necessarily shared” by the department, adding: “Nor do they represent government policy.”