New Home Office guidance on violence without ideology repeats the judgement call that failed three times before Southport. It sets no new duty, no review date, and no way to tell the cases apart.
Estimated reading time: 7 minutes
In short. The Home Office published interim guidance on 27 October 2026 telling frontline staff how to handle "non-ideological extreme violence" (NIEV), violence that is not driven by a clear ideology [1]. It applies only where the person is "not considered at risk of becoming involved in, or supporting, terrorism" (para 2) [1], and states plainly that it "does not create new statutory duties, structures or processes" (para 3) [1]. That terrorism-risk threshold is the same judgement that the independent Southport Prevent Learning Review found had already failed three times, concluding that "this emphasis on ideology created a gap" through which Axel Rudakubana was closed to Prevent without being referred to Channel [4]. The government's own commissioned reviewer, Lord Anderson, considered and rejected a standalone system for exactly this population eighteen months earlier, because telling the two kinds of case apart at referral stage is, in his words, impossible [3]. Three documented mistakes: four stars.
Three times between 2019 and 2021, Rudakubana was referred to Prevent. Three times, the case was assessed and closed. Seven months later, he murdered three children at a dance class in Southport. The independent review into what went wrong did not find that officers broke policy. It found that they followed it, and that the policy asked the wrong question: whether he had an identifiable ideology, rather than whether he was dangerous. On 27 October 2026, the Home Office published interim guidance intended to help the same kind of case get handled better next time. It hands the same question back to practitioners, with no new tool for answering it.
A scope test built on the judgement that failed
The guidance applies "where there are NIEV-related concerns, but the person is not considered at risk of becoming involved in, or supporting, terrorism" (para 2) [1]. Cases that clear that bar go to Prevent. Cases that do not are pointed toward existing safeguarding routes, MAPPA, the Multi-Agency Safeguarding Hub, Children's Social Care, Youth Offending Services, Community Safety (para 9) [1]. The document is explicit about what it is not: "This guidance does not create new statutory duties, structures or processes" (para 3), and "there is no instruction that additional structures should be established" (para 10) [1]. It is also explicitly temporary: "This interim guidance is not intended to pre-judge the outcome of Phase 2 of the Southport Inquiry" (para 5) [1], a phase with no published reporting date. No funding, no mandated structure, no stated review point: the word "interim" carries the whole document's accountability.
The system the government's own reviewer rejected
The Home Secretary commissioned Lord Anderson, the Interim Independent Prevent Commissioner, to examine why Prevent failed both Rudakubana and Sir David Amess's killer. His report, published 17 July 2025, considered a proposal close to what the NIEV guidance now does: a separate structure, sitting outside Prevent, for people whose risk centres on a fascination with extreme violence rather than a clear ideology. He turned it down. "The impossibility of telling, at the stage of a Prevent referral, whether a case belongs in one system rather than the other" was his stated reason, adding that such a system "would introduce an additional burden of having to debate, often on the basis of inadequate knowledge, whether a case is better suited to one mechanism rather than the other" [3]. His formal recommendation instead was that "Prevent should remain open to individuals falling within the existing category of 'fascination with extreme violence or mass casualty attacks'" [3]. The NIEV guidance does not engage with that reasoning or explain why a sorting test he called impossible is now workable.
The overlap is in the guidance's own words
This is not a hypothetical clash. The NIEV guidance's own working definition lists "violence fascination" as a common feature of the cases it covers, and adds that "this behaviour is seen in both non-ideological casework and ideological casework" (para 8) [1]. That is the same population Lord Anderson discussed under the label "fascination with extreme violence or mass casualty attacks" and recommended keeping inside Prevent. Separately, the Home Office's own Key Principles of Prevent, published eight months before the NIEV guidance, already states that "a clear identifiable ideology is not mandatory to make or adopt a Prevent referral" [2], and that the statutory threshold set out in section 36(3) of the Counter-Terrorism and Security Act 2015 asks only whether "there are reasonable grounds to believe that the individual is vulnerable to being drawn into terrorism" [5], with no mention of ideology at all. Both documents already agreed that ideology should not be the deciding factor. The NIEV guidance's own scope test still asks, in substance, which side of a line a case falls on that the guidance's own definition admits is not clean.
The only new instruction is to keep considering
Where that leaves an uncertain case, the guidance says: "Where concerns overlap or Prevent relevance is uncertain, practitioners should still consider whether to make a Prevent referral" (para 16) [1]. That is the same instruction that was already in force when Rudakubana's case was assessed three times. The Southport Prevent Learning Review's verdict on what happened when practitioners made that judgement is direct: "this emphasis on ideology created a gap where a highly vulnerable individual, whose vulnerabilities may have also made him highly susceptible, has been closed to Prevent without consideration for support through Channel" [4]. Fifteen months after that finding, the new guidance offers practitioners the same discretion and no new criteria for exercising it differently.
Three councils did the work nobody has to fund
The guidance showcases what good practice looks like. Northumberland County Council built an enhanced Multi-Agency Safeguarding Hub with a dedicated case management meeting spanning children's and adult social care, health, education and policing. Birmingham's Children's Safeguarding Partnership developed screening tools so practitioners systematically weigh a young person's fascination with violence. Cardiff Council runs a standing multi-agency panel, including Youth Justice and community organisations, for people below the Prevent threshold whose risk has not gone away (paras 10, 11, 16) [1]. All three are named as models. None is required, funded or timetabled for anywhere else. A council without the spare capacity to build its own version gets the same guidance and a different outcome.
Credit where due
The guidance is candid where candour is hard to find elsewhere in this field. A footnote states plainly that "the evidence base relating to NIEV remains emergent and subject to considerable limitations," drawn from "a relatively small number of cases" [1], rather than overstating a thin research base. It also warns practitioners against "single-cause explanations, including those that seek to explain acts of violence solely through mental ill-health or a fascination with violence" (para 7) [1], a more careful line than much of the public debate around Southport has managed. And the three case studies it cites are real, specific and checkable, not generic aspiration dressed up as example.
The claims, tested
| The document's own words | What the record shows | Verdict |
|---|---|---|
| Applies where a person is "not considered at risk of becoming involved in, or supporting, terrorism" (para 2) [1] | Lord Anderson found "the impossibility of telling, at the stage of a Prevent referral, whether a case belongs in one system rather than the other" [3] | The guidance's scope test rests on a judgement the government's own reviewer said cannot reliably be made at this stage |
| "Violence fascination" is a listed NIEV feature, "seen in both non-ideological casework and ideological casework" (para 8) [1] | Lord Anderson's Recommendation 2 keeps exactly this population, "fascination with extreme violence or mass casualty attacks," inside Prevent [3] | The guidance's own definition overlaps with the category its own government's reviewer said should not be carved out |
| "Where concerns overlap or Prevent relevance is uncertain, practitioners should still consider whether to make a Prevent referral" (para 16) [1] | The Southport review found this same "consider" judgement, made three times, left Rudakubana "closed to Prevent without consideration for support through Channel" [4] | No new decision criteria accompany the repeated instruction |
| "This guidance does not create new statutory duties, structures or processes" (para 3) [1] | Three named councils already built the enhanced structures the guidance describes as good practice, on their own initiative [1] | Good practice is shown, not funded or required, leaving it dependent on local capacity |
Verdict
Four stars, from three documented mistakes. This is not guidance that invents a problem or dodges the one it names: it quotes its own evidence limits honestly, warns against the single-cause thinking that dogs this subject, and points to genuine local innovation. What it does not do is close the gap the Southport review actually found. That review did not conclude that practitioners needed a reminder to keep considering Prevent. It concluded that the emphasis on a single question, is there an ideology, had already cost a Prevent referral its proper hearing three times over. The new guidance keeps that question at the centre of its own scope test, offers no sharper way to answer it, and leaves the fix to whichever council has the people and the budget to build one.
The star score counts three documented mistakes: the guidance's scope test turns on whether a person is a terrorism risk, the same triage judgement Lord Anderson's Lessons for Prevent review found impossible to make reliably at referral stage, without engaging that finding; its own definition of non-ideological extreme violence includes "violence fascination" as a feature "seen in both non-ideological casework and ideological casework", the same population Lord Anderson recommended keeping inside Prevent rather than routing elsewhere; and the guidance's only instruction for uncertain or overlapping cases, to "still consider" a Prevent referral, repeats without modification the judgement the Southport Prevent Learning Review found had already failed three times before the attack. Three falls in the 1 to 3 band: four stars; the bands are on the ratings page. This piece makes no finding that the guidance is dishonest or that the frontline practitioners who will use it are at fault; the finding is that the document's central sorting test, and its only answer for hard cases, both predate and do not address evidence already published by the government's own commissioned review. Checked directly against the interim guidance's full HTML text, the Key Principles of Prevent guidance, the Lessons for Prevent review, the Southport Prevent Learning Review and section 36 of the Counter-Terrorism and Security Act 2015 on legislation.gov.uk, all fetched 28 October 2026.
Sources
- Interim guidance to support the management of non-ideological extreme violence concerns (accessible), Home Office, published 27 October 2026
- Key principles of Prevent (accessible), Home Office, published 23 February 2026
- Lessons for Prevent (accessible), Lord Anderson of Ipswich, Independent Prevent Commissioner, published 17 July 2025
- Prevent Learning Review: Axel Muganwa Rudakubana (accessible), Home Office
- Counter-Terrorism and Security Act 2015, section 36, legislation.gov.uk