On November 19 2025, a group of protesters assembled outside France’s premier security event, Milipol, to denounce what they called the salon de la répression – expo of repression – and its welcoming of 38 Israeli defence exhibitors. Barely a week before, French President Emmanuel Macron had ignored his country’s policy since the spring 2024 of excluding Israeli defence companies from French expos, opening the way for their participation in Milipol as a “gesture to facilitate dialogue with Israel”.
I could hear the protesters’ chants as I left the exhibition hall in Villepinte, on the outskirts of Paris. This was my first time attending Milipol and it’s safe to say that I would have been more at home among the protesters than the attendees streaming out of the venue that evening. Unlike the 30 000 other visitors, I wasn’t there to excitedly scope the latest gadgets of the defence industry; I was there to learn about and to document how the industry sells its version of ‘security’.
Space invader
Working for a racial justice organisation, walking into the cavernous conference centre felt like stepping onto another planet – a stifling one, for all its size. I thought of Space Invaders, an arcade game where the goal is to shoot and defeat waves of descending aliens.
Space Invaders recalls the
close relationship between defence and commercial gaming, with modelling and simulation essential to the modern army’s capabilities. The annual
Connections Wargaming Conference has been held every year since 1993, and was
established to “revolutionise how the military trains and rehearses its missions”. Like a child at the arcade, Europe is, almost giddily, bracing to respond to cascading waves of perceived threats. I could feel the excitement of those around me as I walked around the Milipol exhibition hall. At Milipol, I was the alien: an intruder in this space dedicated to glorifying the tools of repression and violence.
One representative for an Israeli facial recognition firm was eager to make his case. “You know what’s happening now in our region,” he said in a low voice. In this context, they’d been able to prove the value of their tools to “find the real bad guys.” He lamented that Belgium, the country where I live, has privacy laws that make it hard to do business with state authorities. “If they don’t want to catch the terrorists, so be it”, he shrugged.
Rah-rah, AI
Enthusiasm for the potential of AI to transform the security sector was on full display at Milipol, seemingly unbothered by
concerns of an impending bursting of the AI bubble. French government representatives announced plans for France to shift from a merely “AI-friendly” country to one that’s “AI-native”, and touted their belief that AI offers opportunities not just to be more efficient, but for true innovation. As an example of this “innovation”, a high-level representative from the Ministry of the Interior argued that the use of AI-powered surveillance at the 2024 Paris Olympics allowed the police to identify situations needing particular attention – to find the “needle in the haystack” – and to “re-centre” its relations with the public and “get closer to users”.
Local groups have a different take:
for them, the deepening of social control and technological surveillance, particularly for working class communities, is the real “achievement” of the Paris Olympics. Those same Olympics gave rise to the
biggest military presence in Paris since World War II, with 10 000 soldiers stationed in a nearby military camp and 45 000 police and gendarmes patrolling the city. The security operation for the Olympics, which also
included anti-drone radars, evolved from France’s Opération Sentinelle, which responded to terrorist attacks a decade before.
A representative from the European Commission underscored for her audience that the drafting of the EU’s AI Act was informed by intensive dialogue with the “repressive sectors” to ensure its drafters were well-equipped with the relevant “know-how”. As it happens, during the political negotiations that followed, France played a key role in the weakening of the original text, including by insisting on loopholes for security and defence, in anticipation of its planned scaled-up surveillance operations during the Olympics.
The forecast calls for war
An important backdrop to the Milipol event is Europe’s current push to rearm, which includes plans for considerable investment in technology.
In March 2025, the European Commission and High Representative presented a
white paper on European Defence Readiness 2030 as part of the ReArm European Plan/Readiness 2030. The White Paper “frames a new approach to defence and identifies new investment needs” and calls for transforming defence through “disruptive innovation and AI integration”. Last October, European heads of state published
Council Conclusions on European defence and security emphasising the need for Europe to “reduce its strategic dependencies, address its critical capability gaps and strengthen the European defence technological and industrial base accordingly”. In other words, rearming Europe goes hand in glove with boosting European security and tech sectors.
The European Investment Bank made its first appearance at Milipol this year, reflecting its unprecedented move to begin financing defence initiatives –
providing support to European businesses through “investment in our industrial base, technological superiority and essential defence infrastructure”. The scope of this support includes dedicated military equipment, infrastructures, services and technologies. Additionally,
a growing number of national investment instruments are actively investing in military and dual-use technologies or announcing their plans to do so.
The push to rearm Europe is accompanied by increased investments in the expansive apparatus of state violence. A representative from Horizon Europe talked up the EU’s investment of 156 million euros under its
Civil Security for Society program, intended to beef up the EU’s responses to a laundry list of threats, including terrorism, organised crime, and border management, and more generally support “security research and innovation”.
Dispelling the clouds
Tellingly, even more present in the exhibition hall than the latest combat gear or weaponry were stands showcasing software. One French drone manufacturer told me that his company was founded by a software engineer. In many instances, the drones, the cameras, the smart patrol cars, were not the main draw; they were accessories to the software that gives them their most attractive features.
The growing emphasis on AI – and in particular general purpose AI – has brought into clearer focus the infrastructure behind it, including the cloud.
Computational infrastructures – meaning the cloud and its accessories – are an important, but less visible, part of the
story of securitisation and tech. French giants Dassault Aviation and Thales
announced a partnership just days after Milipol “for the development of controlled and supervised AI for defence aeronautics”. Dassault Aviation is also a partner, with Airbus and Indra Sistemas, in a project to develop a European combat “system of systems” called the Future Combat Air System: a next generation fighter jet and swarming drones, connected with “combat cloud” technology.
The dominant players in the cloud – Amazon, Google and Microsoft – were not present at Milipol. They are, nonetheless, key players in the infrastructure that underlies state violence, from the Israeli military’s colossal surveillance and targeting of Palestinians using
Microsoft’s Azure, Google Cloud and
Amazon Web Services (AWS), to the US Citizenship and Immigration Services (ICE)’s partnership since 2017 with Palantir, which relies on AWS cloud computing.
According to Brendon O’Connor, “the detention and deportation apparatus that Trump inherited from Obama was built on Amazon” – specifically on its cloud computing service, AWS. In April, NATO inked a deal with Palantir for “intelligence-enabled battlefield operations”. In October 2025, European member states called for efforts to boost Europe’s “modern” defence capabilities, “in full coherence with NATO”.
The cloud allows the embedding of repression within our mobile devices. Geo-location using advertising data captured from mobile phones was, for instance, among the services plugged by companies at Milipol this year, including Israeli firms Rayzone and WaveGuard Technologies, and the Italian company RSC Lab. This technology is
not new, but has gained notoriety thanks to ICE’s deployment of Webloc, which uses location data from mobile phone apps to
“monitor trends of mobile devices that have given data at those locations and how often they have been there”. Webloc was created by
Penlink, a Milipol vendor and the US leader in adtech-based surveillance.
Recent reporting has also revealed ICE’s use of an app called
ELITE (Enhanced Leads Identification & Targeting) that generates, based on data captured from a variety of sources, an interactive map that identifies potential targets for deportation – effectively turning the brutality of immigration enforcement actions into a game of Pokemon Go.
Looking behind the technology – the devices, tools, cameras – reveals not only the material infrastructures underlying them like data centres and undersea cables, which are rightfully receiving growing attention, but also layered
infrastructures of power and dominance. This dominance extends beyond defence, transforming how we interact with
private and public institutions that are increasingly relying on digital interfaces to mediate relations with the public, and everything from
agriculture to
gaming, in ways that can
facilitate or
entrench repressive dynamics.
Events such as Milipol bring home the theatre of security performed by states, institutions and companies. Not only are they a place to sell products, they sell a vision of security that is increasingly mundane, axiomatic, digitised and most of all embedded. These new realities require ever deeper and broader practices of resistance.
