De retour de la biennale des Surveillance Studies
Un compte rendu en anglais de ce que j’ai fait et vu à la dernière biennale du réseau des Surveillance Studies.
I had the chance of attending the 11th Biennial Conference of the Surveillance Studies Network in Lille last month.
I contributed to the collective thinking through two talks that, while distinct in their immediate focus, aimed to untangle the complex relationship between state power, surveillance, and technological infrastructure across the liberal and authoritarian spectrum.
Big Tech, AI and the Military: Shifting Public-Private Configurations
My first talk was part of the roundtable on “(Big) tech as intelligence power: marketisation, capitalism and resistance”, organized by my colleagues Sebastian Larsson, Hajer Ben Jaffel and Emma McCluskey. I focused on the relationship between Big Tech, AI, and the military-industrial complex.
In line with my recent piece in Le Monde diplomatique, my argument centered on the “frenemy” relationship between Anthropic and the Pentagon. It builds on previous research regarding public-private dynamics during post-Snowden controversies to suggest that the public-private boundary in the security field is less a real edge and more an effect of power arrangements (inspired by Tim Mitchell’s Limits of the State, which I keep returning to). In that vein, I tend see the “friction” between the state and a company like Anthropic not as an external confrontation but as internal to a single, internally differentiated security assemblage, where the state retains ultimate coercive capacity, and where the company’s double play is only tolerated as long as it serves the long-term interests of the government.
I also touched upon the necessity for the AI industry to secure Pentagon contracts: with private credit exposure to the sector in the trillions and profit horizons far in the future, long-guaranteed revenue streams from the DoD are again structurally necessary to keep the AI bubble growing. I also built on the work of Elke Schwarz and pointed to the growing accommodation of public procurement to the reduced temporality of tech startups and of venture capital – both in the US and in Europe – which, much like the incorporation of AI to the kill-chain, comes with worrying consequences for military oversight and accountability.
From Blackshirts to ICE: On the History of Political Surveillance Along the Fascism-Liberalism Boundary
My second talk addressed another question that has become increasingly pressing over the past couple of years: is it analytically useful to characterize contemporary techno-authoritarian developments as “fascist,” and to speak of techno-fascism? I argue that it can have heuristic value, but only through a methodological reframing.
Hence, rather than relying on the orthodox historical definition of fascism as a singular historical event, I drew on the landmark work of Alberto Toscano (Late Fascism, 2023), who suggests we treat fascism as a process – a mutable and sometimes contradictory dynamic anchored in colonial capitalism, one that precedes and survives its crystallization in Mussolini’s Italy and Hitler’s Germany.
From this perspective, I explored several structural continuities between the surveillance practices of historical fascist regimes and our contemporary (il)liberal order, namely:
- The colonial origins and the “boomerang effect” (Aimé Césaire): Fascist surveillance did not invent new tools but radicalized a colonial archive of population registries, ethnic classification, and biometric identification. This “boomerang effect” persists today, with technologies like facial recognition field-tested in occupied territories before being sold to European police forces.
- Fascist surveillance builds on liberal surveillance, and vice versa: The massive scaling of political surveillance was not a fascist invention. In Italy, the Central Political Records (CPC) were established under a liberal government in 1894. This process was mirrored in liberal democracies, where the “mimetic rivalry” with fascism led to institutionalizing the “identification and watching” of Communists as a “legitimate” form of statecraft.
- Rule of Law vs. rule by law: A crucial continuity lies in the legal legitimation of repression. Fascist states were not operating outside of legality per se. State violence was largely operated through legal procedures (rule by law) as an instrument of executive power, providing a veneer of legitimacy while emptying out substantive protections. Many of their repressive instruments, such as Italy’s confino (internal exile) or Germany’s Schutzhaft, existed in prior liberal legal codes and were merely expanded and somewhat stripped of judicial oversight – a bypassing of the judicial branch that is also found in recent modes of (il)liberal surveillance.
- Chilling effect as a primary objective: Both historical fascist and contemporary digital surveillance aim not merely at catching dissidents; they produce anticipatory compliance and self-censorship, seeking to preempt dissent from organizing. We see this today in the Chinese state’s focus on suppressing actual mobilizations rather than mere online dissent – a dynamic similar to the one documented in 1930s fascist Italy.
I also highlighted key differences. Among those, the most critical is undoubtedly the automation of surveillance and the way it circumvents the need for widespread social complicity (which was key under historical fascism and today can still go hand in hand with digital surveillance, like in China and arguably other countries). Today, this automation of surveillance has the potential of doing away with the traditional reliance of the nation state on “infrastructural power” – one obtained through indoctrination and/or material rewards for an enrolled avant-garde of the citizenry, performing horizontal surveillance and acting de facto as agents of the state (historically, such infrastructural power has allowed for leverage and forms of popular resistance).
Encounters
This conference brought together critical perspectives on contemporary surveillance. Among those I found most interesting and connected to my own research were Julia Van Dessel’s work on Europol’s Internet Referral Unit and the way it fosters migration-related content moderation on corporate platforms, or that of investigative journalist Crofton Black and his colleagues on phone targeting at global scale, revealing a dataset of 1.5 million lines exposing surveillance of journalists, dissidents, and political figures.
The surveillance and atrocity panel was particularly shocking and moving. Hilla Dayan detailed Israel’s radicalized surveillance tactics, including Unit 8200’s AI tools like “Lavender” and “The Gospel”. She stressed that Microsoft’s contract with Tsahal led to the hosting of some of this data on European Azure servers, which raises huge legal and accountability issues. In the same panel, Dima Samaro demonstrated how surveillance constitutes the infrastructure of occupation, reminding the audience of the facial recognition systems (“Red Wolf/Blue Wolf”) that categorize Palestinians by threat level, humanitarian biometrics used for food distribution in Gaza, and Tsahal’s Genie AI chat-bots producing intelligence on civilians. Lots of emotions too during the screening of Carolina Sánchez Boe’s film, “Digital Detention”, a documentary that exposes the personal stories of people on the move in the United States monitored through invasive GPS monitors and facial recognition apps.
I really enjoyed learning about recent research into Chinese surveillance. Alexander Trauth-Goik for instance analyzed China’s Police Cloud system, showing how 300 million cameras feed into a modular, cybernetic infrastructure influenced by 1960s systems thinking. His analysis of procurement documents helped nuance sometimes caricatural reporting made on state surveillance in China. Other presentations went over the extensive net of privacy regulations and laws developed in China recently as a case of repairing legitimization deficits stemming from “data discontent” towards cybersecurity breaches in mainstream private online service providers.
David Lyon offered a historical corrective, contrasting medieval monastic surveillance – intended for care and salvation – with contemporary practices, suggesting pastoral power’s benevolence offers ethical clues for reforming planetary surveillance. Other contributors included Pete Fussey and Daragh Murray on compound human rights harms, Katharine Hall’s promising archival research on computing’s introduction into policing (1960-1990), and Olga Velikanova on Soviet dossier systems.
Another great conference!