European Radicalisation Is Changing Shape

The new Europol EU TE-SAT 2026 report invites us to examine European radicalisation with precision, responsibility and respect for rights.

The most important finding of the new European Union Terrorism Situation and Trend Report (EU TE-SAT) 2026 is not merely the number of attacks. It is the transformation of the profile of radicalisation in Europe. Europol depicts an environment in which historically recognisable ideologies have not disappeared, but increasingly interact with individual trajectories, personal grievances, psychological vulnerabilities, digital subcultures and forms of violence that do not always possess a coherent doctrinal framework.

This shift requires a change of perspective. It is no longer enough to ask which organisation an individual belongs to or which manifesto they have read. We need to understand how violent identities are formed, which environments legitimise them, which digital mechanisms accelerate them, and which educational or relational gaps enable their consolidation. Read critically, the report does not justify a new season of indiscriminate alarm: rather, it calls for a broader and more demanding assumption of responsibility.

The figures: a high threat that must be interpreted accurately

In 2025, ten Member States reported 45 terrorist attacks: 22 completed, 20 foiled and 3 failed. The completed attacks caused six deaths. In the same year, 486 people were arrested in 21 Member States for terrorism-related offences, up from 449 in 2024 and 426 in 2023. Jihadism accounted for the largest share: 24 attacks and 347 arrests; nine attacks were completed, causing five deaths and 81 injuries. Europol also recorded five right-wing attacks, twelve anarchist and left-wing attacks, and four classified in other or unspecified categories.

These figures cannot be minimised, but neither can they be used without caution. The TE-SAT notes that national laws do not always classify cases in the same way; “violent extremism” has no common legal definition in the Union and is not included in quantitative statistics on attacks and arrests. Furthermore, an arrest is an investigative measure, not a conviction. The report itself warns that the threat level cannot be assessed merely by counting attacks, victims or arrests.

The Italian picture also cautions against reducing the phenomenon to a single strand: Italy reported eleven attacks, all classified as anarchist or left-wing, and 48 arrests distributed among jihadism (26), ethno-nationalism and separatism (13), anarchist and left-wing activity (6), and the right wing (3). Public perceptions of the threat do not always correspond to its actual composition.

From affiliations to trajectories

Europol reports increasingly porous ideological boundaries. Alongside jihadism, right-wing extremism, and anarchist and separatist strands, anti-institutional, accelerationist and nihilist currents are gaining visibility. In some environments, violence is no longer merely an instrument serving a political project: it becomes a language of identity, a means of gaining recognition or an end in itself. Symbols, targets and repertoires can circulate between apparently incompatible ideological worlds.

This fluidity does not mean that ideology has become irrelevant. An interpretation that reduced every radical pathway to a simple individual disorder would be as inadequate as the old search for a single cause. Organisations, narratives and worldviews continue to provide justifications, enemies and models of action. What changes is how they are absorbed: in a more selective, superficial and hybrid form, within biographies marked by frustration, isolation, the search for status or vulnerability.

Prevention must therefore shift from focusing solely on “suspect affiliations” to concrete processes: escalating language, fascination with violence, the breakdown of relationships, compulsive consumption of extreme content, movement from open platforms to closed spaces, and the acquisition of operational capabilities. This is not about monitoring dissent, but about recognising promptly when cognitive radicalisation turns into a willingness to use violence.

The digital sphere is not the background: it is part of the infrastructure

The report emphasises the role of the online environment. Terrorist and extremist content circulates on major platforms as well as smaller forums, audio services, gaming-related spaces and encrypted networks. Borderline content — disinformation, hatred, violent irony, memes and seemingly ambiguous edits — can provide an entry point. Interested users are gradually directed towards more explicit material and more closed communities. In 2025, Member State authorities and Europol assessed 81,862 items of online content linked to terrorism, resulting in 61,412 referrals to 99 hosting service providers.

Artificial intelligence adds another layer: image, audio and video generators have been used to produce propaganda, deepfakes, memes and hateful content; bots and chatbots have been used to expand recruitment, optimise engagement and conduct preliminary research useful in preparing attacks. Much material remains in the realm of the “hateful but lawful”, below the threshold for automatic removal. It is precisely this grey area that makes a policy limited to deleting manifestly illegal content inadequate.

Platforms cannot present themselves as neutral channels. Recommendation systems, gamification, the pursuit of engagement and the persistence of decentralised networks alter the speed and depth of exposure. Their responsibility includes algorithmic transparency, adequate moderation of audio and user-remixed content, independent risk assessment and effective channels of cooperation with authorities, while respecting fundamental rights.

Young people and minors: failure cannot be privatised

The generational finding is perhaps the most alarming: the average age of those arrested was 27, and 130 were eighteen or younger; the youngest was twelve. The presence of minors cannot be explained solely by the availability of propaganda. It raises questions about the quality of educational relationships, the ability to recognise distress, digital loneliness, access to psychological services, and the possibility of finding belonging and recognition in non-violent contexts.

Blaming everything on families would be a convenient form of public abdication. Many parents lack the tools to understand rapidly changing digital environments; schools and local services often intervene only when the pathway is already advanced. Stable multidisciplinary teams, digital literacy, teacher training, proportionate protocols and accessible listening services are needed. Mental health is a vulnerability factor, not an automatic explanation for violence, and certainly not a label to attach to anyone experiencing distress.

Jihadism and the responsibility of Muslim communities

A European Muslim platform cannot ignore the weight of jihadism in the report. Twenty-four attacks out of 45 and 347 arrests out of 486 demand clarity. Denying or minimising this reality would be irresponsible. Yet it would be equally false to turn jihadism into a synonym for Islam or to regard religious practice as an indicator of danger. Of the 398 arrested individuals whose citizenship was known, 259 were Union citizens: the threat cannot be portrayed solely as something arriving from outside, nor addressed by stigmatising entire populations.

Muslim communities assuming responsibility does not amount to collective guilt, nor does it mean accepting the role of religious police on behalf of the state. It means investing in competent theological education, the ability to discuss texts instrumentalised by propaganda, educators familiar with young people’s language, and spaces for dialogue where questions, conflicts and anger can be expressed without being handed over to digital recruiters. It also means recognising signs of isolation and violence at an early stage and building trusted relationships with schools, social services and institutions.

At the same time, public authorities must avoid delegating prevention to mosques as though every jihadist trajectory originated within them. The report shows the opposite: many pathways are individualised, digital and weakly rooted in structured religious knowledge. Communities can be essential partners, but responsibility remains shared and requires mutual trust, the protection of rights and protection from both jihadist and racist, anti-Muslim violence.

From diagnosis to action: five commitments for Europe

The radicalisation described by the TE-SAT 2026 cannot be addressed through isolated responses or measures activated only after an emergency. It requires shared but not undifferentiated responsibility: every actor must intervene within its own remit, with clearly defined expertise, tools and limits.

1. Build permanent public policies

Institutions must move beyond episodic interventions and establish stable prevention systems capable of integrating educational, psychological, social, religious and investigative expertise. Programmes must be evaluated according to their results, not merely the resources spent. Security safeguards must also remain compatible with proportionate interventions, the protection of fundamental rights and the guarantees of the rule of law.

2. Hold digital environments accountable

The responsibility of platforms cannot end with removing manifestly illegal content. The effects of recommendation systems, gamification and borderline content must be understood and made verifiable: even when such content does not constitute explicit propaganda, it can normalise violence and gradually lead users towards extremist environments. Greater transparency, independent risk assessments and effective data access for qualified researchers are required.

3. Strengthen schools, local services and families

Prevention must begin before distress turns into isolation or attraction to violence becomes visible. Schools, social services and families should not be turned into surveillance instruments, but enabled to listen, recognise signs of vulnerability and offer credible alternatives. Digital literacy, critical thinking, psychological support and opportunities for participation are integral parts of a forward-looking security policy.

4. Enhance the role of religious communities and civil society

Religious communities can provide educational and theological authority, challenge interpretations that legitimise violence and build non-confrontational spaces of belonging. Their involvement, however, must not result in the stigmatisation of Muslims or the delegation of state functions. Cooperating in prevention means assuming real responsibilities without accepting collective suspicion as a condition of institutional dialogue.

5. Treat public discourse as part of prevention

Politics and the media also bear direct responsibility. Selecting data to fuel identity-based fears or conflating terrorism, immigration and Islam does not help us understand threats: it creates simplifications, weakens trust and intensifies polarisation. Rigorous reporting must distinguish between different phenomena and convey their complexity, avoiding reinforcement of the very social fractures that extremist actors seek to exploit.

From security to democratic resilience

The TE-SAT 2026 confirms that the terrorist threat in Europe remains high, but above all that its ecology is changing. Lone actors and small self-organised cells, simple methods, unstable ideological identities and easily accessible operational expertise make it harder to detect the transition to action. Twelve of the 22 completed attacks were carried out by lone actors, often using immediately available weapons.

The response cannot consist solely of more control. It requires more knowledge, more coordination and a greater capacity for early intervention. It must also protect the Jewish and Muslim communities identified by the report as targets of threats linked to international conflicts, without importing logics of collective confrontation into European life. Finally, we must distinguish dissent, political radicalism, violent extremism and terrorism, because confusing these categories weakens both freedom and the effectiveness of security.

Assuming responsibility means rejecting two mirror-image alibis: that of those who attribute violence to an entire religion or community, and that of those who, for fear of stigmatisation, avoid confronting the interpretations, networks and language that legitimise violence. A mature European Islam has an interest in embracing this responsibility: not as a collective defendant, but as a civic actor capable of prevention, self-criticism and building common security.

Democratic security does not arise solely from the ability to stop an attack. It arises from the ability to prevent violence from becoming an identity, algorithms from replacing education, and fear from turning citizens and communities into mutual enemies. This is the European responsibility that the Europol report, beyond the statistics, entrusts to us.

Massimo Cozzolino