After Prevent, After Southport

What the inquiry reveals about digital harm, adolescent deterioration, parental blame, and the systems that still fail to intervene meaningfully

See our companion piece by guest author Kat Sumner here

The Southport inquiry’s central finding is stark: this attack could and should have been prevented.

That does not lessen the horror of what happened. It does not soften responsibility. And it does not ask anyone to look away from the suffering of the children, adults and families whose lives were shattered.

But if we are serious about prevention, we also have to resist the false clarity that often follows atrocity. Public discussion narrows very quickly into a search for the person or people who can carry the blame. The child becomes a monster in waiting. The parents become either villains or cowards. Neurodivergence becomes a dark clue. The internet becomes a vague backdrop of evil. That impulse is understandable. It is also one of the ways real lessons are lost.

Some children and adolescents do frightening things. They may carry knives. They may fixate on violence. They may search for disturbing material. They may frighten professionals, schools and parents. But that does not mean their pathway is fixed. It does not mean every child with overlapping traits, diagnoses or referrals is heading towards atrocity. And it does not mean the right response is to widen criminal suspicion while leaving the digital environment itself largely unchanged.

That is why the Southport report matters beyond this one case. It exposes a safeguarding culture that still does not know how to hold youth, neurodiversity, online life, family strain and serious risk together at the same time. It shows what happens when systems oscillate between underreaction and overreaction, between diffuse concern and procedural hand-offs, between blaming parents and missing the architecture that made failure likely in the first place.

One reason the public conversation keeps going wrong is that it reaches too quickly for hard categories. Either a child is evil, or mentally ill, or radicalised, or simply badly parented. But many adolescents who become frightening do not sit cleanly in any one box. They may be spiralling through puberty, exclusion, shame, obsessive online immersion, neurodivergent rigidity, broken belonging and family rupture all at once.

The Southport material itself makes that harder story visible. In year 7 and much of year 8, AR was not yet presenting as the figure people now read backwards from the atrocity. The inquiry records that he initially settled into secondary school reasonably well, with good attendance, generally good behaviour, and a small number of trusted friends. He was quiet rather than overtly alarming. The first clear signs of deterioration came later, towards the end of year 8 and into year 9, when behaviour, conflict and anxiety escalated. That timing matters. It points us back to a period we still do not take seriously enough in safeguarding and education: early adolescence, puberty, social fracture, and the moment when a child’s world can narrow very fast.

That does not mean there was one simple cause. It does mean that deterioration has to be recognised as a developmental process, not just a sequence of isolated incidents. A child can become harder to teach, harder to contain and harder to reach long before adults admit that something is badly wrong. Open internet access is not connection. Endless exposure is not belonging.

That point matters because one of the headline lines from the inquiry has been that he was not suffering from a serious mental illness at the time of the attack. That may be clinically important, but it should not be misread as meaning he was well. The absence of a formal serious mental illness diagnosis is not the same thing as the absence of profound psychological deterioration, relational breakdown or unmet need. The inquiry itself points to a gap in service provision for violence-fixated children and young people who do not have a serious mental health disorder but still need structured psychological intervention.

What stands out most, though, is how little meaningful holding there seems to have been once things began to fall apart. A child should not be left for years to deteriorate with open access to the internet, fragmented adult oversight, and too little meaningful structure, containment or connection. That is not neutral. It is an environment. And in the digital age it is often an accelerant.

This was not only a story of missed referrals. It was also a story of missed ecology. The inquiry is clear that online harms were not incidental. It found that his digital life was repeatedly underexplored, that institutions were too cursory in examining how he was spending his time online, and that there was a direct line between earlier warning signs and later patterns of violent and degrading online consumption. It also found that weak age assurance on X made it probable that he was able to view the Bishop Mar Mari Emmanuel stabbing footage before the attack.

That matters because it shifts the frame. Too often, once a young person becomes frightening, the question becomes: why did the parents not stop it? But the Southport report points to something harder and more honest. Parents were being left to manage a digital environment they did not build, could not fully see, and in many cases could not safely control. The report is critical of the parents, including their failures to set boundaries and disclose what they knew. But it also records a family life shaped by fear, evasion and loss of control. In that context, telling parents simply to put controls on is not a safeguarding strategy. It is an abdication disguised as advice.

That is one of the things we have been arguing through Safe by Default from the beginning. Families are expected to manage multiple apps, passwords, filters, browser settings, consoles, networks and operating systems, while also dealing with vulnerable, traumatised or neurodivergent children who may be impulsive, obsessive, dysregulated, socially isolated or highly motivated to bypass controls. Our campaign has argued that responsibility has to move upstream from isolated parents to the systems, defaults and public services that shape children’s digital lives.

The current conversation about serious youth violence and extremist risk still tends to split in the wrong places. One branch moves towards surveillance, referral and criminalisation. Another falls back on vague calls for awareness, resilience or digital literacy. Neither is enough. The Southport inquiry itself points instead towards a more grounded chain of responsibility: stronger filtering and monitoring in schools, better local authority understanding of online harms, closer attention to evasion tools like VPNs, and more serious institutional curiosity about what children are actually doing online when they are not in education, not engaged and not meaningfully supervised.

That is the space where this debate needs to move. Not just towards more powers, but towards more competence.

Because what was missing here was not only control. It was expertise.

There should be specialist digital safety support for families dealing with high-risk online behaviour. Not a leaflet. Not a generic police warning. Not a school assembly about internet safety. Real, practical, technically competent support for parents who are trying to manage a child whose online life has become frightening, compulsive, exploitative or potentially dangerous. In practice, that means families should be able to access someone who can help lock down devices, identify circumvention, preserve evidence, and judge when an online pattern has shifted from troubling to dangerous. They should be able to get help with router-level controls, parental management systems, browser restrictions, account recovery, secondary accounts, VPN use, reporting routes, and the practical thresholds for escalation. That support should sit before the criminal threshold wherever possible, and alongside safeguarding where necessary.

This is especially important for neurodivergent adolescents. Not because neurodivergence predicts violence. It does not. But because some neurodivergent young people can be more vulnerable to compulsive seeking, literalism, intense interests, humiliation spirals, black-and-white thinking, social naivety, manipulative online ecosystems and professional misreading. Alice Siberry’s recent work is important here. She argues that practitioners often perceive neurodivergent young people, especially autistic boys, as vulnerable to self-radicalisation, but that these perceptions are frequently shaped by surface-level understanding of neurodivergence. She warns that inadequate specialist training can lead to misinterpreting behaviour, mismanaging vulnerability and criminalising neurodivergent young people, while obscuring broader social and technological contexts.

Jacob Astley’s recent work strengthens that case. He and colleagues argue that online extremism and radicalisation are not just about harmful content reaching passive users, but about social worlds, affective dynamics and digital spaces where meaning, belonging and identity are made. That matters because it moves the question away from what did the child view and towards what kind of online and offline world was this young person inhabiting, and how well do our institutions understand it. It also helps explain why a child can be surrounded by content and still starved of meaning, recognition and grounded forms of belonging.

It is also why the Southport lessons should not be reduced to a call for more Prevent. Parliament’s recent report on combatting new forms of extremism warns that Prevent is becoming saturated with cases that are not straightforwardly ideological and recommends a triage function above Prevent so cases involving vulnerability, violence fixation, neurodiversity or mixed and unclear pathways can be directed to the right support. That matters because some young people move through overlapping worlds of grievance, violent spectacle, misogyny, fixation, notoriety-seeking, social breakdown and online immersion that do not map neatly onto one ideology or one intervention pathway.

Part of the problem is that Prevent has increasingly become a holding space for cases that are frightening but poorly understood, where violence fixation, neurodivergent distress, grievance, social breakdown and online immersion are present but ideology is partial, unstable or secondary. That does not mean ideology is irrelevant. It means that referral on its own is not analysis, and it is not intervention. A child can be referred, discussed, flagged and still remain fundamentally unread.

Sandra Ferreira’s recent paper on coercion is useful here too, even though it comes from mental health rather than youth safeguarding. What she adds is a language for how systems can privilege official categories and institutional logics over lived experience, and how the language of risk, control and best interests can obscure what interventions actually feel like and do over time. That is relevant to adolescents who are already frightened, dysregulated or relationally cut off. It reminds us that a system can frame itself as protective while still missing the human meaning of what is happening.

Southport is not the only recent case to expose how serious warning signs can gather without a system knowing what to do with them. In the Tristan Roberts case, recent reporting suggests a different trajectory and a different target, but some recognisable features recur: online rehearsal, weapon acquisition, family fear, misogyny and failed efforts to get help. That does not make these cases the same. It does suggest that the deeper problem is wider than any single label. Some young men move through overlapping worlds of grievance, fixation, humiliation, domination and violent fantasy that do not sit neatly inside one service threshold or one policy category.

The scale of the digital environment is easy to underestimate. Youth Endowment Fund data from 2025 found that more than four in five 13 to 17 year olds had seen online conversations about hurting specific groups, and over a third had taken part in those conversations either to support or challenge them. Mental Health Foundation work also found very high use of online communities among young people, with many describing them as sources of connection and confidence, while also reporting widespread exposure to harmful or disturbing content and gaps in meaningful online safety education. The point is not that the internet is simply toxic. It is that it is both connective and dangerous, and systems need to be adult enough to hold both truths at once.

So yes, public protection matters. Serious warning signs must be taken seriously. Some young people are genuinely dangerous. But a mature prevention model would not begin and end with blame. It would ask harder questions earlier.

Who is helping the parents when ordinary controls are no longer enough?

Who is assessing the digital environment, not just the child?

Who is joining up school filtering, home technology, family fear, neurodiversity, violent preoccupation and service thresholds?

Who is able to act before a family is choosing between silence, surveillance and catastrophe?

At the moment, too often, the answer is no one.

That is why the inquiry’s digital findings should not be treated as a side issue. They are part of the core story. A system that leaves families alone with high-risk digital ecosystems, then blames them after the fact, is not a safeguarding system. It is a liability-transfer machine. A system that notices disturbing online interests but fails to build real support around them is not doing prevention. It is documenting drift. And a system that responds to every alarming adolescent only with either punitive hindsight or diffuse referral will keep missing the middle, which is where real prevention lives.

What would a better response look like?

It would start earlier, around the points where school belonging begins to fracture, attendance starts to wobble and a young person’s world narrows rather than expands. It would treat the secondary transition, exclusion risk and prolonged absence from education as developmental safeguarding concerns, not just behaviour or attendance matters.

It would include a genuine therapeutic pathway for violence-fixated children and young people who do not meet a narrow serious mental illness threshold but are clearly escalating in dangerous ways. It would not leave them with long waits, hand-offs and informal drift. It would ask not only whether a child meets a diagnostic threshold, but whether they are becoming progressively more unreachable, more isolated, more preoccupied with violence, and more cut off from ordinary structures of life.

It would still include serious action where serious danger is present. But it would not stop there. It would build safety by default into devices and platforms used by children. It would require meaningful age assurance and make circumvention harder. It would ensure schools and children’s services are equipped to identify and respond to online harms in a way that is actually developmentally and digitally literate. It would create specialist digital safety support for parents, especially those supporting neurodivergent or vulnerable children. And it would build a triage system above or alongside Prevent, capable of distinguishing between vulnerability, fixation, online immersion, grievance, ideology and escalating violence, rather than collapsing them into one frightening category and hoping referral itself counts as intervention.

The Southport inquiry shows what happens when we fail to do that. It does not tell us that difficult children are doomed. It does not tell us that neurodivergence is a warning sign in itself. It does not tell us that parents are all-powerful. It tells us something both simpler and more demanding.

We have built a digital environment full of violent scripts, weak safeguards and easy evasions. We have left many families to manage that alone. We have not built enough therapeutic or developmental response for adolescents whose lives are visibly narrowing and hardening. And we still do not have enough of the practical, relational and technical support needed when a child begins to spiral.

That is not parental failure.

It is system design.

We are writing this as founders of Safe by Default, a parent- and survivor-led campaign calling for digital safety by design, stronger default protections, better school and service responses to online harm, and access to specialist digital safety support for families, especially those supporting neurodivergent and vulnerable children. More detail is set out on our Safe by Default campaign page.

Further reading

Southport Public Inquiry, Volume 1
Especially Chapter 1 on the inquiry’s core findings and Chapter 6 on online harms, digital access, and weak oversight.

Southport Public Inquiry, Volume 2
Especially the chapters on policing, healthcare, education and the final recommendations. These show how risk was repeatedly identified, passed around and still not meaningfully held.

Home Affairs Committee, Combatting New Forms of Extremism (2026)
Useful on mixed and unclear pathways into violence, online radicalisation, and the need for a triage function above Prevent.

Alice Siberry, written evidence to the Home Affairs Committee (2026)
Important on how neurodivergent young people can be misread through simplistic extremism frameworks.

Extremism and Young People educator brief (2026)
Practical and accessible, with useful material for schools and frontline staff on neurodiversity, online risk and adolescent vulnerability.

Jacob Astley and colleagues on online extremism and educational safeguarding
Useful for understanding online radicalisation as a social and affective world, not just a content problem.

Sandra Ferreira, Held but not healed – Why coercive practices undermine mental health and wellness (2026)
Not about this case directly, but very useful on how systems privilege official categories, risk language and institutional logic over lived experience and human meaning.

Youth Endowment Fund, Children, Violence and Vulnerability 2025
Helpful on the scale of young people’s exposure to harmful online conversations and violent content.

Mental Health Foundation, work on online communities, safety and young people
Useful for a more balanced account of the internet as both connective and dangerous.

Safe by Default campaign page
For our argument on digital safety by design, stronger defaults, and specialist digital safety support for families.