UK: Nearly one in five Prevent referrals had autism, new Home Office figures reveal

Nearly one in five people referred to Britain's controversial Prevent counter-extremism programme had autism recorded as a condition, according to new Home Office figures, raising fresh concerns over the scheme's impact on neurodivergent people.

Analysis of the figures show that 1,833 of the 9,957 people referred to Prevent between October 2024 and September 2025 had autism recorded, with 1,199 classified as confirmed cases and a further 634 recorded as suspected.

The Home Office data also shows that almost 10,000 people were referred to Prevent in England and Wales during the period - the highest annual total number of referrals since records began in 2015, following a 39 percent increase on the previous year.

Overall, 36 percent of the 9,957 individuals referred to Prevent had at least one mental health or neurodiversity condition recorded. Autism was the most commonly recorded neurodiverse condition.

Jacob Smith, Freedom of Expression and Belief team leader at Rights and Security International, told Middle East Eye that the increase in referrals of people with autism was "particularly concerning" and reflected longstanding problems with the government's counter-extremism programme. 

"We've been aware for a while that autistic people are overrepresented within Prevent, but the fact that it has increased so much over the last year is particularly concerning," Smith told Middle East Eye.

"Prevent has, since its inception, been plagued by accusations of bias and discrimination.

"It initially targeted Muslim communities and, although the representation of Muslims within Prevent has changed over time, we still see the disproportionate representation of Asian and Black people. We're now seeing that with autistic people as well."

The figures also showed:

• a 20 percent rise in referrals for individuals related to far-right extremism. 
• Asian people continue to be overrepresented in the Prevent programme, making up 18 percent of referrals where ethnicity was recorded - double their proportion of the wider population in England and Wales.
• Black people are also overrepresented in the Prevent programme - (eight percent of referrals when ethnicity was noted, compared to four percent of the population
• Despite government warnings that “Islamist extremism” remains the greatest threat facing the UK, referrals categorised under that label have declined to eight percent of all Prevent cases.

Smith argued that the programme's referral process relies too heavily on subjective judgement with public sector workers legally obliged to refer anyone to Prevent if they are suspected of terrorism.

"Prevent is set up in a very vague way and relies heavily on referrers' gut instincts," he said.

"That effectively encourages them to take their biases into account when making decisions, whether intentionally or subconsciously. To me, that's a key reason why we're consistently seeing the overrepresentation of certain groups within Prevent."

The Home Office data showed that autism became more prevalent as cases progressed through the Prevent process. Around 23 percent of individuals whose cases were discussed by a Channel panel had autism recorded, rising to 24 percent among those formally adopted into the government's de-radicalisation programme.

The figures form part of the Home Office's second statistical release to include mental health and neurodiversity data following the introduction of the Prevent Case Management Tracker (PCMT) in 2024.

The department said the new system had improved how such conditions are recorded but warned that the quality of the data is still developing.

The Home Office stressed that the figures do not demonstrate a causal relationship between neurodiversity and susceptibility to radicalisation.

"It is important to note that these findings describe associations within the data rather than causal relationships," the report said. "The analysis does not indicate that being neurodiverse or having a particular mental health condition increases or decreases the likelihood of progressing through the Prevent or Channel process."

Criminalising sterotypes about autistic people

But Smith warned that the programme risked reinforcing and criminalising sterotypes about autistic people. 

"A lot of the language used about autistic people [by Prevent] points to traits that are directly linked to autism - for example, hyperfixation. The implication is that autistic people are more likely to pose a risk because of characteristics associated with their diagnosis."

"Instead of addressing the underlying issue, there's a degree of doubling down by relying on autistic people's diagnoses as an explanation," he added.

Following an initial screening and assessment, Prevent referrals which are deemed at risk of radicalisation may be passed to a multi-agency "channel panel". 

These panels are chaired by local authorities and tasked with determining a person’s susceptibility to radicalisation, and whether a tailored package of support is necessary to address the risk.

Of the 1,100 referrals taken on as a channel case in the year to September 2025 where ethnicity was known, 835 (76 percent) were white, 159 (14 percent) were Asian, 59 (five percent) were black and 47 (four percent) were "other".

Smith said people referred to Prevent could face lasting consequences even when no further action is taken.

"As we documented in our 2024 report Caught in the Web, information about Prevent referrals - including referrals that don't go anywhere or are mistaken or malicious - is shared with a wide range of policing bodies, as well as other public authorities, including the security services and immigration authorities," he said.

"While much of what happens within Prevent is secret, there is a real risk of long-term consequences for people who are referred, not only in their interactions with the police but also in their interactions with other public services."