Counter-Terrorism Law and Free Speech on Campus

A rare Equality Act 2010 (EqA) challenge to a referral under Prevent, the UK’s counter-terrorism programme for identifying and responding to suspected signs of radicalisation, has ended with a County Court finding that a school unlawfully discriminated against a Muslim schoolboy. Although a first-instance decision, the ruling has important implications for universities. As well as foregrounding a risk inherent in Prevent’s precautionary referral process that the lawful expression of staff and students may be recast as evidence of radicalisation through a linguistic gloss unsupported by the underlying facts, it points to a potentially underused route to legal redress where such reinterpretations result in discriminatory treatment because of religion or protected philosophical belief.

The County Court Case

The case concerned a 13-year-old Muslim pupil who was involved with several other children in a playground incident. Teachers concluded that he had encouraged another pupil to place a note containing a divisive message on the ground and had shouted words from an internet meme referencing Islamic terrorism. A referral marked “Prevent” was sent by the school to the local authority safeguarding hub, which had the capacity to make an onward referral.

The court found that the language of the referral went beyond the facts established by the school’s investigation, with the teacher having cumulatively “glossed” the incident using religiously loaded language. That language was “a powerful indicator that the teacher’s subconscious thought processes were influenced by his knowledge that the boy was a practising Muslim”. Although there was no finding that the teacher had consciously intended to discriminate, the court concluded that the pupil had been subjected to unlawful direct discrimination because of his religion.

How Prevent Referrals Work

Understanding the significance of the ruling – and why “Prevent” is not just a verb in the infinitive form, or a “referral” simply a matter of sending someone to the Headteacher – requires a detour through the finer points of the UK’s counter-terrorism law.

Prevent is one of the four strands of CONTEST, the Government’s counter-terrorism strategy. Described by the Home Office as an “early intervention programme” operating in the “non-criminal space”, Prevent operated for years as a policy before being placed on a statutory footing by section 26 of the Counter-Terrorism and Security Act 2015. That provision requires specified public authorities, including schools and universities, to have “due regard to the need to prevent people from being drawn into terrorism”.

The programme’s first objective is to tackle what the statutory guidance calls the “ideological causes of terrorism”, including by reducing the so-called “permissive environments” in which radicalising narratives are thought to flourish. The remit is deliberately broad, with the guidance identifying Islamist and extreme right-wing terrorism as the principal domestic threats, but also referring to conspiracy theories as possible “gateways” to radicalised thinking and to other terrorist narratives involving antisemitism, misogyny, “anti-establishment” or “anti-LGBT grievances”.

How does this work in practice? The guidance envisages a “notice, check, share” process, in which frontline professionals raise concerns regarding possible radicalisation internally, usually with a designated safeguarding lead or Prevent lead. An external referral may then be made to specialist police officers, who determine whether the individual should be considered for support through Channel, a voluntary, multi-agency programme. Other cases may be handled through a Police-Led Partnership.

As the policy language surrounding Prevent suggests, there is a form of behaviourism at work here, with exposure to particular ideological narratives understood as setting people on a “pathway” towards terrorism, and early intervention presented as a way to interrupt that process. Having conceived the subject as what mid-twentieth-century sociologists would have termed a “cultural dope”, the programme tends to regard ideological stimuli as producing progressively reinforcing effects that draw people toward increasingly extreme material, with insufficient allowance for their cognitive ability to question, resist or reinterpret what they encounter, whether immediately or as their thinking develops over time.

As a result, otherwise lawful expression or conduct is forever at risk of being over-interpreted, not because of what it means in the here-and-now, but because of what it hints at for tomorrow. And since the interpretative world of those who make referrals to the Prevent programme is no less susceptible to contemporary prejudices and assumptions than anyone else’s, these subjective judgments can find their way into an official counter-terrorism concern, with potentially far-reaching consequences for those whose identities or beliefs render them vulnerable to suspicion.

That is no exaggeration, either, since even where a police-received referral is judged not to warrant further intervention, it can leave a documentary footprint on the Prevent Case Management Tracker (PCMT), a national database maintained by counter-terrorism policing, where information is ordinarily retained for six years after case closure. Referral data is also frequently stored in local safeguarding databases – including in schools, local authorities and NHS infrastructure – where policies may allow retention far beyond the PCMT baseline. There are now multiple documented cases in which individuals referred to Prevent – despite not meeting the threshold for further action – suffered serious detriment simply because their names were retained in the system.

The Inference of Discrimination

At first glance, the County Court’s finding – that the pupil’s religion could have influenced the teacher’s reasoning without his consciously intending to discriminate – might seem to involve a far-reaching principle with implications for freedom of speech, permitting a court to infer from a person’s words and the surrounding evidence that considerations they themselves deny were relevant nevertheless influenced their reasoning. Indeed, scepticism about the very idea of subconscious reasons was expressed nearly 30 years ago by Lord Browne-Wilkinson, who argued in his dissent in Nagarajan v London Regional Transport that a consideration not consciously taken into account could not properly constitute a “reason” for acting. But the principle is not an innovation of this ruling: the law of direct discrimination has long recognised that a protected characteristic may materially influence less favourable treatment without the decision-maker consciously intending to discriminate.

In this case, although the facts – including the language of the referral – gave rise only to a prima facie inference of discrimination, the Trust lacked the contemporaneous evidence needed to rebut it. Under section 136 of the EqA, where a claimant establishes facts from which a court could decide, in the absence of another explanation, that discrimination occurred, the burden passes to the respondent to show that it did not. In Igen Ltd v Wong, for instance, the Court of Appeal held that, because the facts necessary to explain the treatment will normally be in the respondent’s possession, a court will ordinarily expect “cogent evidence” once that burden has shifted.

Applying those principles, the judge held that the Trust was unable to provide such evidence that its treatment of the pupil was “in no sense whatsoever” related to his religion. So while the court did not find “smoking gun” evidence of a discriminatory mindset, it was able to rule that the school had discriminated against the pupil because the violent and religious overtones of the referral gave rise to an inference of discrimination that the Trust could not rebut – largely because it lacked contemporaneous notes and records showing how the decision to refer had been made.

Finding that the treatment amounted to direct discrimination because of the protected characteristic of religion under section 13, the judge awarded the pupil £10,000 for injury to feelings.

Prevent and Freedom of Speech on Campus

Although the ruling is a non-binding first-instance County Court decision, its potential application to higher education is difficult to dismiss, not least because the EqA applies similar provisions to schools and universities.

Section 85 prohibits the responsible body of a school from discriminating against a pupil, including by subjecting the pupil to “any other detriment”, while section 13 defines direct discrimination as less favourable treatment because of a protected characteristic.

For universities, section 91 uses almost identical language, prohibiting the responsible body of a university from discriminating against a student, again including “by subjecting the student to any other detriment”. If a student’s religion or race significantly influences the way their words or conduct are interpreted, described or presented in a Prevent referral, the same question therefore arises under section 13: whether the student was subjected to less favourable treatment because of a protected characteristic. University employees are likewise protected against discriminatory detriment under section 39(2)(d).

But the implications extend beyond religious discrimination. Under section 10, the protected characteristic of “religion or belief” also encompasses qualifying philosophical beliefs, including gender-critical beliefs and certain anti-Zionist beliefs, both of which have featured prominently in recent university free-speech disputes. A university could therefore find itself facing rather awkward questions if a student’s protected philosophical beliefs were recast in a Prevent referral as evidence of anti-LGBT, extreme-right or Islamist radicalisation, particularly where the language of the referral went beyond what the student’s original words or conduct, or other contemporaneous evidence, could reasonably support.

Whether it’s desirable that heterodox academics and students should find themselves relying on discrimination law, with its established routes to judicial redress, rather than dedicated statutory protections for freedom of speech and academic freedom, whose individual enforcement mechanisms are considerably more limited, is perhaps a question for another day. But the fact that this question arises at all suggests that universities need as many effective safeguards for lawful expression as they can get.

It’s true that the statutory Prevent duty guidance: for England and Wales makes clear that specified authorities, including universities, must continue to comply with their other legal obligations when carrying out the duty, “particularly” those arising under the EqA. That guidance also states that each authority should maintain appropriate records demonstrating compliance with Prevent and, specifically, should consider whether its record-keeping arrangements for Prevent referrals are “proportionate and appropriate”.

It’s also true that universities are subject to an additional statutory protection. While section 26 of the Counter-Terrorism and Security Act 2015 sets out the core Prevent duty, section 31 requires universities, when carrying out that duty, to have “particular regard” to their statutory duty to secure freedom of speech and to the importance of academic freedom. As Ouseley J established in R (Butt) v Secretary of State for the Home Department [2017], the section 26 Prevent duty does not override the section 31 free-speech safeguard. Moreover, the Court of Appeal in Butt [2019] endorsed his further conclusion that “non-violent extremism which carries no risk of drawing people into terrorism is not subject to the guidance”.

To this apparently formidable array of free-speech safeguards, the Office for Students (OfS) adds its own advice, urging English universities to record decisions likely to have a substantial effect on lawful speech, demonstrate how they have had particular regard to its importance, and preserve those records so that they remain available for external review.

And yet, the difficulty lies in the deliberately precautionary nature of the referral system itself.

On the one hand, the Prevent duty guidance says there should be “one consistent and proportionate threshold… across all extremist ideologies and radicalisation concerns”, and that someone making a referral should be able to “indicate why the risk is relevant to Prevent”. On the other hand, frontline professionals are asked to consider whether someone “may be on a pathway that could lead to terrorism”, and told: “it is far better to receive referrals which turn out not to be of concern than for someone who genuinely needs support to be missed”.

In effect, the system is asymmetric about error. A false negative may have exceptionally serious consequences, while a degree of false-positive referral is tolerated in the interests of safeguarding. The operation of such a system on campus creates an obvious risk not only that the section 26 duty may eclipse the section 31 safeguard, but that the language used to turn an ambiguous concern into a Prevent referral may itself raise questions of direct discrimination under the EqA.

The Documentary Trail

That makes the process by which somebody becomes a “concern” in the first place exceptionally important. The guidance envisages two distinct stages before an external referral is made: universities should have a process for staff or students to “raise concerns regarding radicalisation internally”, followed by a process through which those concerns “are then assessed”.

But the existence of a two-stage process does not necessarily guard against discriminatory assumptions, since a student’s religion or protected philosophical beliefs – and the assumptions made about them by those reporting or assessing a concern – could influence how their words or conduct are interpreted at either stage. A lecturer, erring on the side of caution, might accurately report what a student has said, only for an internal assessment to place an unwarranted religious or racial gloss on it. Alternatively, that gloss might enter with the initial report and subsequently be adopted by the university’s Prevent lead.

The County Court ruling is particularly instructive here, with the judgment turning in part on the disparity between the limited contemporaneous evidence of what the school’s investigation had established and the more elaborate account subsequently presented in its Prevent referral.

For universities, the implication must surely be that the documentary trail cuts both ways. A proper contemporaneous record may help demonstrate why a concern was raised and how it was assessed, but the language of that record may also provide evidence of the assumptions through which otherwise ambiguous conduct came to be understood as a Prevent concern. And where the burden of proof subsequently shifts under section 136 of the EqA, an institution unable to reconstruct its reasoning may face the converse problem: an absence of cogent evidence with which to show that a protected characteristic played no part in the treatment.

The Scale of the Problem

In a sense, statistics are the wrong register for assessing referral decisions which, if taken incorrectly, can draw someone into a counter-terrorism surveillance assemblage whose reach into their life may have profoundly damaging consequences.

Nevertheless, OfS figures published in June 2026 show that 250 potential radicalisation cases reached an institutional Prevent lead in 2024–25, down slightly from 265 the previous year. Yet formal external referrals rose sharply, from 65 to 95 – an increase of 46 per cent and the highest number recorded since the current data series began in 2020–21. Nor was the increase confined to a handful of institutions: 55 higher education providers made at least one formal referral, up from 39 the previous year.

Over the past decade, a series of incidents has shown how readily lawful academic work can be drawn into the hinterland of this surveillance assemblage: students warned to act with caution when engaging with an essay by a leading left-wing academic on the ethics of a socialist revolution; discussions about referring a student’s artwork on the geopolitics of the Middle East to Prevent teams; a criminology lecturer reporting that she had run her course reading list past the police because of concerns about its content; and a Muslim PhD student cancelling a conference on counterterrorism legislation citing fears of Prevent.

More recently, six students at the University of Essex were referred to Prevent after their Palestine Solidarity Society shared social-media posts from Middle East Eye and Al Jazeera concerning the death of Hamas leader Ismail Haniyeh. Although Essex Police dropped its investigation into the students for allegedly “supporting a proscribed group”, the university’s disciplinary case continued for more than 200 days before being abandoned. One of the students, a black Arab Muslim woman, later said she believed she had been “specifically targeted and disproportionately affected” by the process.

The recent County Court ruling suggests that the legal question in cases such as these may not simply be whether there was a proper basis for making a Prevent referral, but how the underlying conduct was interpreted and subsequently presented in that referral: what was actually said or done, what was added in its retelling, and how far the resulting interpretation had travelled from the underlying facts. Where that interpretative gap reveals the influence of a protected characteristic – including religion or belief – the EqA may provide a route to legal redress for individuals whose lawful expression has been recast as evidence of extremism.