TEARDOWN Published 15 November 2026 at 08:12. Evidence-based. Source-cited. No sponsored content.

Prevent referrals hit a record high this year, with fewer cases than ever tied to a clear extremist ideology. The same watchdog report shows the Home Office left its own prevention budget unspent for six years, then cut funding to the highest-risk areas.

4 out of 5 stars4/52 documented mistakes in this teardownHow ratings work

Estimated reading time: 6 minutes

The National Audit Office building on Buckingham Palace Road, Victoria, London, seen from street level against a clear sky.
The National Audit Office, Buckingham Palace Road, Victoria, London, 22 July 2014. Photo: Danrok / Wikimedia Commons, CC BY-SA 4.0.

In short. The National Audit Office's report on the Prevent programme, published 13 November 2026, records 10,293 referrals in the year to September 2025, "an all-time high" [1], page 17, with only 30% now showing "a clear, dominant extremist ideology," down from 73% in 2017 [1], page 22. The same report finds that "rising referral volumes" are "placing growing pressure on the system" and that the Home Office "has not done enough to understand or respond to these pressures, jeopardising the sustainability of the programme" [1], page 14. Its own financial analysis shows the department underspent its Prevent budget in every year from 2020-21 to 2025-26, by 28% in the latest year alone, while cutting real-terms priority funding to the highest-risk local authorities by 54% over the same period [1], pages 10 and 26 to 27. The NAO's own conclusion: "There is little justification for the level of underspending Prevent has seen in recent years" [1], page 14. Two documented mistakes: four stars.

Prevent is a small programme by Whitehall standards. The Home Office's own internal budget for it in 2025-26 was £23.4 million, against a counter-terrorism system that costs more than £3 billion a year and attacks whose indirect economic costs alone the department has put at £4.11 billion for five incidents in 2017 [1], page 26. What the NAO's report adds is not a judgement on whether Prevent works. It is a set of figures, drawn from the Home Office's own records, that sit uneasily next to the department's own account of a system under growing strain.

A record year, and a thinner picture of risk

Referrals to Prevent fell to a low of 5,416 in the year to September 2020, during pandemic restrictions, and have risen every year since, reaching 10,293 in the year to September 2025, "the highest number of referrals ever received in a single year" [1], page 4. The Home Office has not formally analysed why, but its working view is that awareness of Prevent rose after the July 2024 Southport attack [1], page 17. At the same time, the referrals themselves have become harder to categorise. In the year to September 2025, only 30% presented "a clear, dominant extremist ideology," against 73% in the year to September 2017, an average decline of 5.3 percentage points a year [1], page 22. In 2024-25, cases with no identified ideology became the single largest referral category, 2,859 of them, against an average of 697 a year between 2019-20 and 2023-24 [1], page 22.

The NAO's summary draws the two trends together directly: "rising referral volumes and an increasing proportion of referrals without a clear, dominant extremist ideology, are placing growing pressure on the system. Workloads are increasing with more borderline cases being discussed by panels, increasing the risk of inconsistent decision-making between local authorities" [1], page 14. That is the department's own evidence base for saying Prevent needs closer attention, more resource and better national oversight, not a critic's gloss on it.

The budget that sat unused

Set against that picture, the report's account of the Home Office's own spending is direct. "In 2025-26, the Home Office had a budget for Prevent of £23.4 million. Of this, it spent £16.8 million, an underspend of 28%" [1], page 26. That was not a one-off. "The Home Office underspent its Prevent budget in every year from 2020-21 to 2025-26, by an average of £5.6 million (17%), in real terms," a total of £33.7 million left unspent over the six years [1], page 27. The department told the NAO that the underspends were "partly attributable to weaknesses in financial decision-making and capabilities": accruals processed incorrectly, inaccurate forecasting, VAT double-counted in places, budgets set against the wrong codes [1], page 26. The underspend narrowed in the two years to 2024-25 before widening again in 2025-26, which the Home Office attributed to structural changes within the department [1], page 26.

None of those explanations describe a programme short of money relative to its own stated budget. They describe a department that, in the same years its own data shows referral volumes climbing and ideology-clarity falling, did not spend the comparatively modest sum it had already set aside for the job. The NAO's own assessment does not soften that: "There is little justification for the level of underspending Prevent has seen in recent years, in a context in which the Home Office acknowledges local workload pressures and has delayed some of its own research" [1], page 14.

The funding cut to the areas rated highest risk

The central underspend sits alongside a separate reduction, this time a deliberate one, in funding to the local authorities Prevent itself identifies as facing the greatest risk. "The Home Office has consistently underspent its own Prevent budget while reducing its funding to local authorities." Priority area funding, targeted at areas the department's own assessment rates highest risk, fell from £9.5 million across 45 local authorities in 2020-21 to £4.4 million across 27 local authorities in 2026-27, in real terms, a reduction of 54% [1], page 10. The department's stated reason was "partly in response to the Home Office's interpretation of performance monitoring data, which found no notable difference in performance between funded and non-funded local authorities" [1], page 10.

That justification is about relative performance between areas, not about whether local capacity itself is under pressure. The same report records that Channel panels, the multi-agency bodies that assess the hardest cases, are already absorbing more "borderline cases," with "increasing the risk of inconsistent decision-making between local authorities" named as a direct consequence [1], page 14. A fund built to support the areas facing the most risk was cut by more than half in real terms during the exact period the report says that risk, and the workload that comes with it, was rising.

The claims, tested

The report's own words What its own figures show Verdict
"Rising referral volumes and an increasing proportion of referrals without a clear, dominant extremist ideology, are placing growing pressure on the system... jeopardising the sustainability of the programme" [1], page 14 The Home Office underspent its own £23.4 million Prevent budget by 28% in 2025-26, continuing a six-year pattern that left £33.7 million unspent in total [1], pages 26 to 27 A department that says pressure is mounting left money it had already allocated to the programme unspent, every year, through the period it says that pressure built
Priority funding to high-risk local authorities was cut because performance monitoring "found no notable difference in performance between funded and non-funded local authorities" [1], page 10 The same report records growing caseloads and "the risk of inconsistent decision-making between local authorities" as a live, worsening problem [1], page 14 Funding to the areas rated highest risk was cut by 54% in real terms while the report's own evidence shows local strain increasing, not easing

Credit where due

The NAO's report does not treat Prevent as a failing programme, and is specific about what works. It found "evidence that the programme has a positive impact on those who receive support," and that "Channel panels demonstrate strong multi-agency collaboration and accountability" [1], page 14, based on direct observation of six panels in operation [1], page 14. The Home Office was also candid with its own auditors about the specific, mundane causes of its underspend, incorrect accruals, VAT double-counted, the wrong budget codes used, rather than offering a vaguer account that would have been harder to check [1], page 26. And the report is explicit that Prevent "is not an expensive programme, particularly compared with the economic and human costs of a terrorist attack," a fair caveat on the scale of the sums involved [1], page 14.

Verdict

Four stars, from two documented mistakes. Gareth Davies, the NAO's head, put the context plainly: "Prevent is operating in an increasingly complex environment, with record referrals and changing patterns of risk" [2]. A department that describes its own programme in those terms, and whose own figures back that description up, should be spending what it has already budgeted for it. Instead the money sat unused for six straight years, and the specific fund meant to back the areas carrying the most risk was cut in half. The NAO's single recommendation on this point asks the Home Office to "continue its work to strengthen its financial management, to avoid significant underspends on its Prevent budget in future" [1], page 14, with no date attached for when that work should be done.

The star score counts two documented mistakes. First, the Home Office underspent its own Prevent budget in every year from 2020-21 to 2025-26, by 28% in 2025-26 alone and £33.7 million in total, during the same years its own referral data shows volumes reaching an all-time high and the proportion of cases with a clear extremist ideology falling from 73% to 30%; the NAO found "little justification" for it. Second, real-terms priority funding to the local authorities the Home Office itself rates highest risk was cut by 54%, from £9.5 million across 45 authorities to £4.4 million across 27, over the same period the report records growing caseloads and a rising risk of inconsistent decision-making between local authorities. Two falls in the 1 to 3 band: four stars; the bands are on the ratings page. This piece makes no finding about whether Prevent's underlying casework, thresholds or Channel panel model are effective, about the conduct of any named official, or about the Home Office's wider counter-terrorism budget beyond the Prevent programme figures the NAO itself audited; none of those were independently tested here. The source was fetched and read in full directly on 15 November 2026.

Sources

  1. Reducing radicalisation: the Prevent programme, National Audit Office, published 13 November 2026
  2. Stronger Home Office oversight needed as Prevent referrals hit record high, National Audit Office press release, 13 November 2026
  3. File:National Audit Office building, Buckingham Palace Road.JPG, Wikimedia Commons, photo by Danrok, 22 July 2014
  4. Creative Commons Attribution-ShareAlike 4.0 International licence
SHARE THIS ARTICLEXBlueskyMastodonLinkedInRedditEmail