
More autistic people referred to Prevent, new Home Office statistics indicate
6th August 2026
Eighteen per cent of all Prevent referrals involve people with autism diagnoses or ‘suspected’ autism, the Home Office announced today in its annual Prevent statistical release, covering the year from October 2024 to September 2025. This figure is up from the 14 per cent indicated in the last statistical release and equates to 1,833 autistic or potentially autistic people.
‘This is a shocking number of autistic people in England and Wales who now have long-lasting police records under a secretive programme that doesn’t require any evidence that someone is committing a crime’, said Jacob Smith, Team Leader on Freedom of Expression and Belief at Rights & Security International (RSI), which has investigated Prevent and Channel for over a decade. ‘There is no excuse for unleashing this much suspicion against autistic people.’
The increased referrals of autistic people and those with ‘suspected’ autism come at a time when the Home Office has admitted that autism does not cause violence. RSI has published a literature review reaching the same conclusion: there is no persuasive evidence of a link between autism and violence.
This increasing of Prevent referrals that target autistic people is coupled with what RSI sees as worrying expansion in the overall quantity of referrals: for the period covered by the release, there were 10,293 referrals — amounting to a 39 per cent increase. Despite this increase, authorities ultimately decided that 79 per cent of referrals required no further action, or diverted them outside of Prevent to other services.
Outside of these headline changes, many of the statistics paint a picture that is similar to previous releases:
- People of Asian descent are significantly overrepresented in Prevent (18 per cent, when ethnicity was noted), in comparison to their representation among the population of England and Wales (9 per cent). (The Prevent statistics released today cover England and Wales; statistics on Scotland are released separately, and the Prevent duty does not apply in Northern Ireland.)
- People of Black descent are also overrepresented (8 per cent of referrals when ethnicity was noted, compared to 4 per cent of the population).
- Referrals involving a person with no identified ideology continue to make up the majority of cases (56 per cent), with only 7 per cent of those cases adopted at the later, more intensive ‘Channel’ stage.
- The proportion of referrals relating to what the government calls ‘Islamist extremism’ has continued to decrease (now standing at 8 per cent) despite repeated government pronouncements that ‘Islamism’ remains the greatest threat to the UK.
- There were, again, 16 referrals involving Northern Ireland ‘Dissident Republican Extremism’ and an unknown quantity relating to ‘Unionist/Loyalist Extremism’, despite the Prevent duty not applying in Northern Ireland.
There is significant overlap between these released figures (covering October 2024 to September 2025) and the previous statistical release (covering April 2024 to March 2025). It is unclear to us why the Home Office has released this data with such significant overlap, which makes analysis more difficult.
More from RSI on this new data soon.
