New government documents obtained by Rights & Security International (RSI) show what we believe is an incoherent and underinformed approach among officials toward the referral of autistic people to the Prevent and Channel counter-terrorism programmes. These documents also point to incoherent and, in our view, careless approaches to assessing Prevent’s other equality impacts, particularly where race and religion are concerned. These deeply flawed practices – showing a seeming indifference to potential ableism, racism, Islamophobia and other discrimination – have persisted for years.
A 2021 report on autism and Channel commissioned by the government, along with several equality impact assessments (EIAs) – all of which were previously unpublished – together create a picture of a government that is aware that autistic people may be over-referred to Prevent and Channel, but does not appear to be taking action. RSI has received other documents of concern that it will publish soon.
The 2021 report explicitly declined to say that autism contributes to ‘radicalisation’.
Prevent is the UK government’s flagship programme for stopping people from being drawn into terrorism, and has long been the subject of critiques from civil liberties and community groups for alleged discrimination and other violations of human rights. Channel is a more intensive follow-on programme from Prevent and has faced similar critiques.
RSI obtained a report on Channel and autism that the Home Office commissioned from a market research firm then known as Ipsos MORI (now Ipsos) and that was completed in 2021. We also obtained three EIAs from 2023 that pertain to Prevent, Channel, and the Home Office’s response to an external review of Prevent completed by Sir William Shawcross in February 2023.
The government has not explained why it has not previously published the EIAs.
The Home Office has suggested in a review of the handling of Southport murderer Axel Rudakubana under Prevent and Channel, prior to his crime, that autistic characteristics and ‘special interest[s] … frequently are combined’ and may make an individual ‘susceptible to being drawn into terrorism’.
RSI is concerned that such a statement does not appear to be supported by the research the government commissioned, or by any kind of consensus among academic researchers on violence. Such language may effectively draw a target on autistic children when it comes to counter-terrorism policing, based on assumptions and stereotypes rather than any professional expertise in autism or consultations with autistic people.
Although disability is a protected characteristic under the Equality Act 2010, and although the 2021 report would have put the government on notice of a potential problem, the EIAs on Prevent and Channel do not provide any statistical or other factual information on the impact of these counter-terrorism programmes on autistic or other disabled people. (For the purposes of the Equality Act, autism may qualify as a disability, although many autistic people may not define themselves as disabled.)
Each EIA mentions autism explicitly – unlike any other disability – and the current Prevent referral form asks specifically for information about neurodiversity. Additionally, the 2021 report the Home Office commissioned on autism and Channel raised concerns, based on a small sample, that up to 26 percent of people referred to Channel might be autistic. (The Financial Times previously reported this figure but did not publish the 2021 document.) However, the government apparently is not collecting systemic data about the impact of Prevent or Channel on autistic or other disabled people.
The Home Office’s approach to equality monitoring does not appear to be any better when it comes to assessing Prevent’s other equality impacts, particularly in relation to the strategy’s impacts on people with specific racial/ethnic identities or people with specific religious beliefs.
In sum, the Home Office is not carrying out the data capture and analysis needed to evaluate whether referrals to Prevent and Channel involve discrimination. Instead, the Home Office appears to assume that because (it says) the programmes are not intended to target any particular groups, then they cannot be discriminatory in practice. This methodology goes against best practices in assessing discrimination, which require the Home Office to conduct a deeper, more rigorous analysis than it is presently doing.
Where race is concerned, other data we have obtained and are publishing today suggest that the situation of inadequate data collection is growing worse: the proportion of Prevent referrals for which the person’s race is not recorded, or is marked as ‘unknown’, has climbed to 60 percent.
The EIAs also raise broader questions about the focus and purpose of Prevent. For example, when discussing the strategy’s impact on people based on their sexual orientation, the Home Office says ‘there is the chance that [Prevent] may improve relations in tackling radicalising ideologies that perpetuate anti-LGBTQ+ narratives’. It is unclear to us why the government seems to take the view that preventing ‘Islamist’ violence should be part of a specific strategy, but preventing violence against LGBTQ+ people can be left to ‘chance’.
For more information, see the below documents:
Home-Office-commissioned report on ASD and Prevent:
1. Ipsos MORI, 'Autism Spectrum Condition: Support within the Channel process’ (November 2021)
Home Office equality impact assessments:
2. ‘Prevent Channel Duty Guidance Refresh’ (July 2023)
3. Prevent Duty Guidance Refresh’ (May 2023)
4. ‘The Response to the Independent Review of Prevent’ (February 2023)
Ethnicity of people referred to Prevent, delimited by outcome categories:
5. NPCC, ‘Response to FOI 181/2024’ (18 February 2025)
RSI’s analysis of the available data: