The Prevent Counter-Terrorism Strategy After Southport

SUMMARY

• The 2024 Southport attacker was referred three times to the counter-terrorism Prevent Strategy without being engaged by the programme, because he was deemed not to be susceptible to terrorist ideologies. This replicated the experience of the 2021 murders in Plymouth.
• The Anderson review, by David Anderson KC, suggests that existing legislation should be amended to include violence-fascinated individuals who demonstrate unclear, or even no apparent, ideological motivation.
• The review also suggests a reorganisation of Prevent into a broader safeguarding and violence prevention scheme.
• This significantly contradicts the thrust of the earlier, controversial 2023 Shawcross Independent Review of Prevent, which mandated the need for a clear and consistent ideological element to be present.

The counter-terrorism Prevent Strategy has recently undergone a further government review by David Anderson KC following the 2024 murders by Axel Rudakubana of three girls in Southport, with the attacker having previously been referred three times to Prevent without being engaged because he was deemed not to be susceptible to terrorist ideologies. This replicated the experience of the 2021 murders of five people by Jake Davison in Plymouth, murders also conducted by an attacker proposed for referral to Prevent, but who was similarly deemed not to be a terrorism risk.

The 2025 Anderson review of Prevent was commissioned to analyse lessons for Prevent from both the Southport murders and the murder of David Amess MP by a man who had previously undergone Prevent counselling.

This article focuses on the issues of the remit and boundaries of Prevent—who can and should be engaged by the programme.

The evolution of the Prevent Strategy

Initiated in 2006 in the wake of the 7/7 London attacks, Prevent is one of the four pillars of the CONTEST counter-terrorism strategy. Over twenty years, the actual focus and content of Prevent have changed significantly. Initially it focused explicitly and exclusively on British Muslim communities, prompting allegations that Prevent in this form was Islamophobic and contradictory to the parallel policy priority of community cohesion.

The coalition government introduced a radically altered ‘Prevent two’ phase. The local community-based activity arm was greatly reduced, while Prevent's remit was widened to all forms of terrorist ideologies. Foregrounded instead was the Channel process, whereby individuals perceived to be vulnerable to terrorist ideology and involvement were identified, referred, assessed and, if suitable, offered preventative mentoring and counselling, with the introduction of a ‘Prevent duty’ for all public servants in education, health and social services.

The nature of Prevent referrals is the key issue addressed by the recent Anderson review. Initially, ‘Islamist’ referrals were the largest category, but this has since been steadily reducing in comparison to ‘extreme right-wing’ referrals. The Shawcross review identified that the overwhelming majority of actual terrorist attacks and late-stage plots in the UK over the past twenty years have been Islamist in nature, so suggested that Prevent was not focussing well enough on its core purpose of preventing terrorism.

Shawcross suggested that many extreme right-wing referrals to Prevent were not serious and implied without evidence that public servants were hesitating to refer Muslims through fear of being branded racist.

The most significant growth towards becoming the biggest Prevent referral category by 2023 was of the ‘mixed, unstable and unclear’ (MUU) category, defined as cases ‘where the ideology presented involves a combination of elements from multiple ideologies (mixed), shifts between different ideologies (unstable), or where the individual does not present a coherent ideology yet may still be vulnerable to being drawn into terrorism (unclear)’. The independent reviewer of terrorism legislation, Jonathan Hall KC, in his own report on Southport, identifies that ‘this unsatisfactory classification was introduced as part of the Prevent triage system in 2017–18 and has since been replaced’.

MUU has subsequently been disassembled into a number of distinct Prevent referral categories, including ‘no ideology identified’ (the largest single category of the most recent annual, 2024–2025, Prevent referrals report). This category attracted significant criticism from Shawcross, who said that it led to many unnecessary referralsand denied it could lead to terrorist attacks. This claim rests significantly on the definition of ‘terrorism’. ‘InCels’ are not currently accepted as being terrorism.

Growing international research evidence questions the extent to which consistent and identifiable (single) ideologies can be identified in many contemporary terrorist plotters and attackers. Just a year after the Shawcross report, Ken McCallum, the Head of MI5, said:

‘Straightforward labels like “Islamist terrorism” or “extreme right wing” don't fully reflect the dizzying range of beliefs and ideologies we see. We're encountering more volatile would-be terrorists with only a tenuous grasp of the ideologies they profess to follow… It's harder… to quickly and definitively determine whether an act of violence is ideologically motivated or driven by another factor like mental health.”

Should Prevent be engaging with violence-fascinated individuals’ (VFIs)? If not, should there indeed be a new and distinct scheme targeting violence-fixated individuals apparently lacking any clear ideological motivations?

The Southport murders and the implications

A public inquiry into the Southport murders and the aftermath of public disorder is currently underway and is due to report early in 2026.

Axel Rudakubana was referred three times to Prevent by schools at different stages of his educational journey, but his case was rejected three times by police without reaching Channel for possible adoption.

The first referral came whilst Rudakubana was at a pupil referral unit and followed him admitting that he had taken a knife to his former school on numerous occasions with the intention of stabbing someone, accompanied by his internet searches around school massacres, with this category used in the initial Prevent referral.

Prevent police judged ‘no CT (counter-terrorism) concern’ present, with mental health-related interventions going ahead instead and leading to a later autism diagnosis. The same school referred Rudakubana a second time, fifteen months later, for online interest in Libya and Muammar Gaddafi, but Prevent quickly judged it again to be of no counter-terrorism concern. A third referral came soon afterwards when the school observed Rudakubana reading online about the London Bridge Islamist terrorist attacks, the Manchester Arena attack, Israel/Palestine and the IRA. Despite this, Prevent quickly judged there to be no vulnerability to radicalisation or associated ideological grievance present, seeing Rudakubana's vulnerabilities as being addressed by his education and healthcare plan. This case arguably echoes that of the Plymouth murderer, Jake Davison. Davison's mother raised concerns about his extreme misogynistic views and he was considered for referral to Prevent, but there was ultimately not a formal referral.

‘Lone actor’ cases like this suggest that Prevent's working understanding for referral and intervention—and of how this relates to other ‘violence-fascinated individuals’—needs to be urgently examined.

What was the Anderson review?

David Anderson's review was to examine the interactions the Southport and Southend attackers both had with Prevent, and to suggest reform. What is clear, as Anderson identifies, is that Prevent is working and can reach such vulnerable individuals. The barrier stopping Rudakubana, Davison and possibly others from being adopted by Prevent has been the insistence that the individual referred must have a clear, recognisable and consistent ideological motivation present. The 2019 guidance from the Prevent national leads confirmed that this was not essential for Channel consideration and adoption, but Anderson identifies that this guidance has often not been followed within the Channel process.

Anderson has recommended that the government should amend the existing Prevent legislation to include violence-fascinated individuals’ who demonstrate unclear, or even no apparent, ideological motivation, since frontline practitioners are already referring vulnerable individuals on this basis. The reasons for this are: the strong overlap of the profiles and modus operandi of terrorists and VFIs; their similar negative impacts on the public; and Prevent's proven ability to identify the latter. Above all, Anderson identifies that the individuals concerned in both categories are ‘lone operators, where the boundary between terrorist and non-terrorist extreme violence is imprecise and mobile’. Establishing a separate, standalone system would be difficult and dangerous. This in itself is a significant retort to the 2023 Shawcross Prevent review and to the police's ground-level implementation of Prevent.

Anderson made a longer-term proposal to integrate the current standalone structure of Prevent within a wider ‘safeguarding and violence prevention strategy’ behind ‘a big front door’ of the local authority's multiagency safeguarding board.

This would ensure that the vulnerabilities and motivations of individuals referred would be considered holistically with effective handovers, whilst genuinely inappropriate referrals of individuals needing other types of intervention would no longer be made directly to Prevent. The integration would bring an extra advantage: a reduction in the public visibility and distinctiveness of Prevent and the potential reduction in associated stigma and controversy over a referral, as a referral would then be a ‘safeguarding’ one, rather than be a ‘Prevent referral’.