Prevention After Berlin
On the evening of 25 July 2026, a 21-year-old German citizen of Lebanese background drove a van into the closing rally of Berlin's Christopher Street Day near Potsdamer Platz and then went after people on foot with a machete. A woman was killed. Twenty-nine people were injured, several critically. Abdul Ballout was shot by police the following day. Germany's interior minister classified the attack as an Islamic-inspired terrorist attack, and the investigation has treated it as such since.
Ballout was not a stranger to the German security services. He had been convicted of preparing a serious act of subversive violence after a failed attempt to travel to join the Islamic State, and he was held in pre-trial detention until May 2026. He was released roughly two months before he drove into the CSD crowd. On his release, a Berlin court ordered him to attend a deradicalisation programme for high-risk offenders; he was assessed twice by the Berlin-based Violence Prevention Network, whose director Thomas Mücke supervised the case. VPN concluded he was not reachable, that his composed and formulaic answers about distancing himself from ISIS were the tell of someone gaming the process, and that there were "no access points" left. VPN documented that finding to the authorities. Ballout was not returned to custody or reassessed. The system did not fail at seeing him. It failed at what to do with what it saw.
Known attackers, known targets
That combination, an attacker already known to the state and released back into the environment that had radicalised him, is the recurring shape of Islamic-inspired terrorist attacks in Europe over the last decade. Anis Amri, who drove a truck into a Berlin Christmas market in December 2016 and killed twelve, had been under German intelligence surveillance since 2015, with active tracking discontinued three months before the attack and deportation stalled across fourteen identities. Usman Khan, who killed two at Fishmongers' Hall near London Bridge in November 2019, had been convicted of a 2010 Islamic-inspired terror plot, released on licence in December 2018 subject to the UK's Desistance and Disengagement Programme, and attacked at a Learning Together prisoner-rehabilitation conference he had been approved to travel to. Kujtim Fejzulai, who killed four in Vienna in November 2020, had been released early after eight of twenty-two months on a court-supervised deradicalisation programme he had, as the Austrian interior minister acknowledged publicly, fooled.
Norway's own experience is similar. On 25 June 2022, Zaniar Matapour opened fire outside the London Pub in central Oslo, hours before the city's Pride parade was due to move through the same streets. Two people were killed. Nine were seriously wounded. Matapour was sentenced in July 2024 to thirty years' detention. Court testimony described an audio bay'ah pledge sent to the Islamic State via Arfan Bhatti, a Norwegian citizen born in Oslo who was in Pakistan at the time and whom an Oslo court convicted in January 2026 of orchestrating the attack; Bhatti has appealed. Both men were part of a small, well-known cluster of Norwegian Islamic-inspired extremists that had been on the security service's radar for roughly two decades before the attack.
The accelerationist far right targets the same population. Three and a half months after Oslo, a 19-year-old accelerationist gunman opened fire at the Tepláreň bar on Zámocká Street in Bratislava. Two people were killed. He published a manifesto in the accelerationist idiom that runs from Christchurch through Bærum and Buffalo to Halle, a lineage tracing to Anders Behring Breivik's 2011 attacks. Breivik's own document was primarily an Islamophobic counter-jihad text; the anti-LGBT+ hostility in it sits within a broader anti-"cultural Marxism" frame that bundled anti-feminism, multiculturalism, and queer rights as symptoms of a Europe made vulnerable to Islamic conquest. The accelerationist writers who came after him have carried the anti-LGBT+ thread forward.
The two traditions are less ideologically opposed than they look. They speak in different surface ideological vocabularies, but both inherit shared Abrahamic scriptural and moral frameworks about gender, sexuality, and civilisational order. They converge on the same target and on similar operational forms: canonical bodies of text whose earlier attackers are cited by name and elevated as reference figures, cross-language propaganda infrastructure, and marginalised, Telegram-native young-male recruitment. The similarities matter more than the differences. Revontulet's applied work is anchored on the safety of the population being targeted rather than on the ideology or origin of the actors targeting them; the Bratislava Pride threat picked up in 2024 is one instance of that approach doing what it is meant to do.
The political response has gone in a different direction. Across Europe this summer, the queer community targeted in Berlin has been used to justify collective punishment of Muslim citizens. Parts of the right-populist and Christian-conservative ecosystem that spent June arguing that queer visibility had become excessive now argue that queer safety demands a defund-and-restrict response against Muslim faith communities. Those two arguments occupy the same discursive space and take turns being loud.
A common accompaniment is the claim that European public discourse suffers from what we in Norway call berøringsangst, a fear of touching the topic of Islamic-inspired violent extremism openly; the empirical case for that at the public-sphere level is weak, given that each Islamic-inspired terrorist attack in Europe generates months of sustained, cross-party debate. A narrower and separate claim, made from within Muslim communities themselves, is worth engaging on its own terms. AP politician Hashim Abdi has called for an internal reckoning inside Muslim communities against radical Islamic-inspired extremism and anti-queer attitudes, noting that individuals who raise the issue face harassment and social control, and warning at the same time that far-right extremists converge with radical Islamic-inspired extremists in their opposition to queer people while weaponising Islamic-inspired attacks to spread further hate against Muslims. A broader version of the reluctance he describes runs across the political spectrum, because confronting religiously-rooted homophobia is uncomfortable across faith traditions, wherever the community expects deference. If the goal is to reduce homophobia and anti-queer violence, the intervention has to target homophobia and anti-queer violence directly, wherever they are rooted. Using the queer community as pretext to escalate against one faith community leaves the broader pattern intact.
The defund-and-restrict template
On 29 July 2026, four days after the Berlin attack and a week after Norway marked the fifteenth anniversary of Breivik's 22 July 2011 attacks, Unge Venstre's leader, Omar Svendsen-Yagci, called for cutting parts of Norwegian state support for Muslim faith communities, roughly 1.3 billion kroner over the last five years, unless those communities meet a policy test that includes support for same-sex marriage. Faruk Terzic, leader of Muslimsk Dialognettverk, rejected the proposal in the same coverage, noting that the network had convened dialogue between religious leaders and queer communities ahead of Norwegian Pride in 2022 and that "one should not punish those who can be important collaboration partners in society."
Europe has already tried the template. Austria's Operation Luxor, launched seven days after the November 2020 Vienna attack, raided sixty sites and arrested thirty Muslim activists and academics; the Graz court subsequently ruled the operation illegal, and the operation produced no convictions. A parallel effort to criminalise "political Islam" was blocked by the Austrian Constitutional Court. France's 2021 loi confortant le respect des principes de la République took the legislative route: the state has since closed 672 Muslim establishments by its own count and dissolved organisations including the Collective against Islamophobia in France, drawing wide criticism across religious and civil-society lines. Neither operation produced a strong empirical counter-terrorism case. Austria's independent Zerbes commission that investigated the Vienna attack found that Austrian domestic intelligence had prepared further measures against the attacker but did not implement them because the service was preoccupied with preparing for Operation Luxor. In France, no publicly documented case ties a closure under the separatism law to a specific foiled attack, and available assessments have found that closed mosques were not the primary radicalisation sites for perpetrators of French terror attacks.
First, the defund-and-restrict template reads radicalisation backwards. Carving out a separate legal and material status for one faith community deepens the perception of that community as second-class members of the polity. That perception is one of the reliable accelerants of the radicalisation the policy claims to address. Ballout's own biography fits that pattern: an earlier extremism-related conviction, a custodial cycle, and a documented history of prior crimes. A policy that adds to that pattern is not counter-terrorism in any operational sense.
Second, visible institutional infrastructure is where prevention has any reach at all, and the template removes it. Funded, visible mosques are on average where moderation happens, and where the state has any chance of seeing when moderation fails. The alternative is what German and French counter-terrorism services have documented under the label "basement mosques": informal, unregistered spaces that fill the gap when the visible institutional ones lose legitimacy, and that recur in the biographies of European Islamic-inspired attackers from the Molenbeek networks around Khalid Zerkani that fed the Bataclan cell onward. The template shrinks the visible surface and grows the underground one.
Third, a principled and secular approach to justice, one that includes both equality of citizens before the law and the prohibition on collective punishment, is exactly the sort of thing a liberal democracy should be protecting, and exactly what the template erodes. Singling out one faith community for a policy test that is not applied to the others violates the equality principle. Holding a whole community accountable for actions committed by individuals who happen to share its religion violates the collective-punishment principle. No connection between Norwegian Muslim communities and the Berlin attack has been reported. Christian denominations in Norway that decline same-sex marriage receive state support on the same statutory basis. Anti-queer positions have Christian, Jewish, and Muslim scriptural traditions that theologically conservative wings of each faith continue to cite, and show up across those faiths in opposition to same-sex marriage, in conversion therapy, in the ostracism of queer congregants, and in everyday discrimination. Reformist and queer-affirming currents exist inside all of them: the Church of Norway has performed same-sex weddings since 2017; Salam, a Norwegian queer-Muslim organisation, was declared bankrupt in March 2026 after losing state funding in successive rounds and struggling through internal disputes; Berlin has the Ibn Rushd-Goethe Mosque, a liberal mosque where LGBTQ worshippers are welcomed. Applying the test consistently across faiths in a way that recognises this spectrum is what a principled and secular approach requires; applying it selectively collapses the argument for secular equality into an argument about a specific target.
What prevention actually looks like
Real prevention is a set of specific, doable things, and the defund-and-restrict template is not one of them.
It looks like actually acting on what deradicalisation workers, probation officers, and intelligence analysts already know, and resourcing the transitions between them so they can act rather than merely track.
It looks like continued and increased support for the imams, mosque leaderships, and civil-society organisations that make up the visible, oversight-reachable parts of European religious life across faiths.
It looks like sustained dialogue in and around that infrastructure: interfaith dialogue, structured contact between government and law enforcement and faith communities, and coordinated work with the civil-society organisations that already defend queer rights and queer safety.
It looks like prevention that starts from what you are trying to protect rather than from which kind of extremism you have decided to oppose. A Pride event facing an Islamic-inspired threat and a Pride event facing a far-right accelerationist threat get the same protective work from the same team.
And it looks like a consistent public standard that the state does not collectively punish a population for what individuals sharing their religion or ethnicity have done. Norwegian Muslim families whose community centres would lose funding had nothing to do with the Berlin attack. The prohibition on collective punishment is one of the things a counter-terrorism policy is meant to protect. A policy that abandons it in the name of protection has failed at its own stated purpose.
Revontulet's work is intelligence: analysis and network mapping that surface threats early enough to act on them. We start from the evidence rather than from a preferred set of adversaries, treat threats as network problems rather than sorting problems by ideology or ethnicity, prioritise preventive work over repressive tools and civilian-side infrastructure over military and surveillance frames, take diversity and pluralism as premises on the observation that violent extremism is the violent denial of diversity and the durable antidote is more pluralism, and hold to equality before the law as a value a serious safety operation defends rather than trades away. The values of a free society and the operational architecture of protecting it are, in the end, the same object.
Twenty-nine people are recovering, at the time of writing, from injuries sustained in the attack. A woman is dead. Preventing the next attack that fits the same pattern is a real operational problem, and it is worth reaching for the tools that actually work.