How Mandelson reviewer failed to probe counter-terror police failings on Southport killings

The murder of three young girls at a Taylor Swift dance class in Southport in July 2024 shocked Britain.

Alice da Silva Aguiar, nine, Bebe King, six, and Elsie Dot Stancombe, seven, were murdered. The perpetrator tried to kill eight other children and two others, including the class instructor.

Speculation online about the killer’s racial and religious identity triggered a spate of serious rioting in cities across the country.

Just a month after being elected, the new Labour government was knocked off course. 

Because the perpetrator, 17-year-old Axel Rudakubana, did not enter a plea and the court originally entered a not guilty plea on his behalf, facts associated with the attack were not aired in public, allowing misinformation to spread and spiral. 

Last Monday, nearly two years after the attack, a report following Sir Adrian Fulford’s nine-week inquiry into the Southport killings was published.

Fulford, a retired high court judge, was then in the headlines again last Friday when he was appointed to chair a review commissioned by Downing Street into Peter Mandelson’s vetting process.

It was revealed the day before that security officials decided Mandelson, a former Labour peer and ambassador to the US, should not receive developed vetting clearance to become ambassador, but were overruled by the Foreign Office.

Fulford’s report on Southport - more than 750 pages long - found that the killer, referred to as AR, “could and should” have been stopped if his parents and various state agencies had acted sooner.

It transpired during the inquiry that AR was referred to Prevent, the government’s counter-extremism programme, three times - beginning in October 2019 - with no further action taken.

Fulford’s core finding was that no state agency took responsibility for managing AR and his needs. 

In response to a request for comment, the Home Office pointed us to Home Secretary Shabana Mahmood's response to the report on Monday.

She said: “There was a failure by the agencies involved to take responsibility and nobody was clear who was in charge. 

“So the failure, because it belonged to everyone, belonged to no one.”

But is this really true?

In this piece, we examine Fulford’s report. We argue that it failed to properly scrutinise the failings of counter terrorism police after AR was reported to Prevent.

We show that counter-terrorism officers failed to follow the proper process for managing AR’s Prevent referral, keeping his case away from local authority oversight with significant consequences.

As such, many recent criticisms of Prevent have been misplaced, but they have been used to push for a greater role for counter-terrorism police in the Prevent process.

The Home Office has argued that AR was not believed to have a clear ideology despite displaying what Fulford described as “a deep and enduring preoccupation with extreme violence", and so was ignored by Prevent.

But we argue there is no evidence that this was the issue at hand.

Fulford’s report recommends that the Home Office establish a “single agency or structure” to “record, monitor and coordinate interventions for children and young people who present a high risk of serious harm”.

We show that this was already in place. 

Victim of its own methodology

Fulford’s inquiry tells a concerning story that unfolds over five years, with AR’s behaviour becoming increasingly erratic and disturbed.

According to Fulford, his parents bear significant moral responsibility as the day of the killings approached. He gathered weapons in the house and booked taxis for what his father suspected was a plan for some kind of attack.

AR had been subject to multiple interventions by different public agencies. All of them downplayed the risk he posed just as the evidence was suggesting it should be escalated.

Shockingly, AR was discharged from the oversight of the Child and Adolescent Mental Health Services (CAMHS) family support team just six days before the attack.

The report is not a comfortable read. Only a few individuals stand out as having had a proper understanding of the risk AR posed. Their concerns were downplayed by more senior professionals or the agencies to which they reported them.

While the information held by different state agencies was incomplete, much of it was sufficient to justify serious action.

For all its qualities of depth and detail, however, Fulford’s report leaves some serious questions unanswered.

It has become a victim of its own methodology, with media coverage over the past week mesmerised by the failure of any state agency to take ownership of the risk AR posed.

Fulford is sardonic throughout the report about the lack of professional curiosity on the part of some officials.

But the inquiry itself appears to have also lacked curiosity about the precise nature of the arrangements in place for managing Prevent referrals.

Yet understanding this is central to explaining how AR really slipped through the net.

Dovetail unexamined

Dovetail was the name of a multi-agency hub designed to handle Prevent referrals by bringing counter-terrorism police officers and representatives from other agencies (social workers, education and mental health services) together under a local authority coordinator. 

A number of witnesses to the inquiry referred to “Prevent/Dovetail”, while Fulford referred to “Channel/Dovetail” without seriously examining what that signified.

Dovetail has been piloted in a number of areas, including Lancashire, since 2016. After an evaluation of the pilots in 2018, it was extended to other areas and put on a regional basis with “North West Dovetail”, where AR lived, the first area of implementation. 

Andy Burnham, the mayor of Greater Manchester, also requested Dovetail following the inquiry into the 2017 Manchester Arena Bombing, which found that the bomber, Salman Abedi, should have been the subject of a Prevent referral.

But information about Dovetail is extremely hard to come by for ordinary people. 

A July 2025 report titled “Lessons for Prevent” by the government’s previous Prevent commissioner, Lord David Anderson, said Dovetail was closed in March 2024 after being evaluated in 2021, but it did not explain why.

The Dovetail process is as follows: after someone is reported to Prevent, the referral is first assessed to see whether there is an active counter-terrorism interest in the case.

Next, it proceeds to the Prevent Gateway Assessment stage. Here, a review takes place, which determines whether the referral has met the threshold for further consideration. This is supposed to be completed within five days.

If the case is assessed to meet the threshold, it goes onto Dovetail, the multi-agency hub. 

Keeping Southport killer out of Dovetail

The Southport killer was discharged from Prevent early on in the process, supposedly at the Gateway stage, not just once but three times. 

On his first Prevent referral, AR was held at the Gateway stage not for five days but for over six weeks. 

During this lengthy period, a Police-Led Process (PLP) of information-gathering took place - which was highly unconventional, since the PLP is usually invoked only later on, after a panel has already judged that there is a reasonable risk the individual is at risk of radicalisation.

Counter-terrorism officers, therefore, appeared to be keeping AR out of Dovetail.

Anderson’s 2025 report on Prevent noted: “The case was referred to the PLP rather than to Channel, which it was speculated within the organisational learning process may have been influenced by a reluctance to proceed to information-gathering because of local-authority control of the process under Dovetail, or by wrongly applying the section 36 test at the Gateway Assessment.” 

The Section 36 test refers to a stage of the referral process that is supposed to occur only after the Gateway assessment to determine whether someone should be offered support through the Channel de-radicalisation programme. 

It should be made either by Dovetail or by a Channel panel. Fulford’s report noted that invoking a police-led process was unconventional. 

But while his inquiry addressed the separate responsibilities of different agencies and their possible failures, it ignored how counter terrorism police did not follow the correct Dovetail process. 

Failure to act on Prevent referral

Consider the sequence of events.

AR was first referred to Prevent on 5 December 2019 by The Acorns School, a specialist school for children with behavioural difficulties, following a number of incidents which had concerned teachers since his arrival there in October.

Then, on 11 December, separately, he was arrested for an attack at Range High School, his previous school which had expelled him. 

AR went there with a knife and a hockey stick apparently seeking to kill someone he believed had bullied him. Merseyside Police were called and he attacked another pupil with the hockey stick on the grounds that he did not want to be “arrested for nothing”.
 
After the arrest there were two processes underway: one involving Prevent, the other a criminal process. 

On 13 December, the Fixed Intelligence Management Unit and Joint Assessment Team began their initial assessment of AR’s case and passed it on for the Gateway assessment on 16 December, suggesting they saw a clear risk.

What should have happened next was a swift review - taking no longer than five days - to determine that AR’s referral met the threshold for further consideration.

On 17 December, the Mersey Care NHS Trust set up a Multi-Agency Safeguarding Hub (MASH) meeting. 

It was attended by representatives from Merseyside Police (the investigating force), Lancashire police, North West Counter Terrorism Police, CAMHS, Range High School and The Acorns School, and by the designated counter-terrorism officer responsible for the Gateway assessment.

On 23 December, the counter-terrorism officer made an entry into the police case management system saying AR should be referred to the Vulnerability Support Hub (VSH).

This is a Counter Terrorism Policing-run programme (now known as the Counter Terrorism Clinical Consultancy Service) in which people referred to Prevent are assessed by NHS mental health clinicians.

This was highly unconventional.

Fulford’s report said the officer “was right at the time to mark AR’s case as one that required input from the Vulnerability Support Hub”.

But in other documentation provided to the inquiry, it is clear that it was understood that the role of Dovetail included “carrying out the key functions of receiving referrals, assessing vulnerability, supporting panels and raising awareness".

There was no need for the VSH to provide a view at this stage in order to determine that the referral should have gone to Dovetail.

On 3 January 2020, the designated counter-terrorism officer visited AR at home under the Police-Led Process (PLP) arrangement.

AR was also being considered by another Multi-Agency Safeguarding Hub set up in relation to his criminal offence. It had its last strategy meeting on 6 January, which was also attended by the counter-terrorism officer. 

At the meeting, AR’s case was closed by Child and Adolescent Mental Health Services (CAMHS). 

On 19 January, AR was formally charged over the Range High School incident and provided with a 10-month intervention by Lancashire’s Child and Youth Justice Service.

However, AR’s Prevent referral was then closed on 31 January 2020.

This was a full 57 days after it was received, and the case had not even been considered under Dovetail.

Friction with local authority

It is difficult to avoid the conclusion that there was a deliberate effort on the part of counter-terrorism police to keep AR out of Dovetail, which would have put his case under local authority oversight.

This failure to act on the first Prevent referral was highly significant and served as an alibi for others: if Prevent - which is designed to assess the risk of harm to others - discharged AR, then that was seen to indicate that it was only his own vulnerabilities which needed to be addressed.

The report of the Independent Review of Prevent, published before the Southport attack in February 2023, had a telling description of Dovetail. 

Sir William Shawcross, the lead reviewer, indicated that there was friction between counter-terrorism officers and those representing health and social services.

He wrote: “Some within the police felt that the counter-terrorism risk was not being properly managed, as information gathered by the local authority was not being shared early enough in the process.

“There was also concern that placing greater risk management with local authorities has led to weaknesses in the system for mitigating counter-terrorism risk.”

Shawcross further said: “Some practitioners noted concern that referrals might slip through the net and not be passed onto the police, and some police felt that the Dovetail model frustrated their ability to gather information and interact with individual cases.” 

Recall the 2025 report by Anderson, then-Prevent reviewer, who suggested that in AR’s case, there may have been “a reluctance to proceed to information-gathering because of local-authority control of the process under Dovetail”.

In response to the Fulford report, Sir Peter Fahy - a former national lead for Prevent - commented that since the Southport attack, Prevent had become so inundated with referrals that counter-terrorism police were at risk of being overwhelmed.

“They were at risk of becoming social workers,” he said. Yet it was precisely that problem which Dovetail was designed to solve.

Home Office narrative unconvincing

It has been widely suggested that Prevent failed because it was too focused on ideology and AR did not exhibit a clear ideology.

Then-Home Secretary Yvette Cooper said in 2025 that it needed to be established what changes would have to be made to ensure that serious cases involving "mixed or unclear ideology" were not missed.

Her successor, Mahmood, made the same argument last week. But this is unconvincing. 

In June 2019, a joint letter from the Home Office and Counter Terrorism Policing Headquarters was sent to Prevent officers stating that they should pay special attention to referrals with unclear, mixed or unstable ideologies. 

The relevant officers said during the inquiry that they were aware of the letter.

Even if not all the relevant information had been shared between Merseyside Police, the school and Prevent, there was still sufficient evidence for a positive Gateway assessment. All indications suggest that would have been the expected outcome.

So why did the Fulford inquiry not treat Dovetail as more significant?

Why were witnesses not questioned in detail about it?

The heads of Prevent at the Homeland Security Group and North West Counter Terrorism were questioned as part of the inquiry. But they provided standard answers that suggested there was nothing amiss.

Why was the lead for Lancashire Dovetail during the period not called?

The only witness from Lancashire County Council to be questioned on the issue was its director of Public Health, Wellbeing and Communities.

This shows a lack of detailed scrutiny by the inquiry, when compared to the detail of its investigations of other agencies.

The Southport inquiry declined to comment.

An extraordinary misunderstanding

Responding to the report last week, Home Secretary Mahmood said Fulford was “clear that police should have progressed the perpetrator to the multi-agency Channel programme. Channel could have actively assessed and managed his risk.

“Instead, he was not deemed suitable because he had no fixed ideology.”

This statement suggests an extraordinary misunderstanding of the problem. 

Mahmood said that in the second part of the inquiry, still to come, Fulford “will provide recommendations on the adequacy of the existing arrangements across all arms of the state for identifying and managing the risk posed by violence-fixated individuals.  

“He will explore what specific interventions are required to reduce the risk to the public.”

She announced the appointment of a new independent Prevent commissioner, Tim Jacques, a former deputy assistant commissioner for Counter Terrorism Policing and senior national coordinator for Prevent and Pursue.

Policy analyst Clive Walker has described the government’s counter-extremism strategy as being in a “policy spiral”.

This involves “a policy which lacks clear initial purpose or subsequent direction, progression, control and reflection”.

A policy spiral is “susceptible to unresolved contradictions or gaps, dramatic direction changes, and uncertain outcomes”.

One of the consequences of a policy spiral is that implementation of the policy becomes increasingly opaque, even to those working within it.

They may have a grasp of their role in enacting the policy, but not how their part fits with other parts.

Fulford’s report describes a policy spiral and its tragic and - with hindsight - preventable consequences. The risk is that it has simply provided for another loop of the spiral.

The children who died, who were injured or traumatised in the attack, and the parents mourning their loss, deserve more.