This paper examines the relationship between artificial intelligence and the privatisation of war, based on the hypothesis that the rise of AI systems applied to national security does not merely represent a technological evolution, but is a symptom of a more profound transformation: the gradual transfer of functions traditionally carried out by the state — strategic decision-making, surveillance and military targeting — into the hands of private actors. The first chapter establishes the theoretical and historical framework of the thesis. After defining artificial intelligence as a socio-technical construct, it traces the collapse of the Westphalian paradigm — accelerated by the attacks of 11 September 2001 — which paved the way for the transition from warfare based on armies and territories to warfare waged in the information domain. It then analyses the ‘grey areas’ of international law (self-defence, nuclear weapons, ‘below-the-threshold’ operations), which allow hostile operations to be conducted without crossing the formal threshold of an armed attack, and the relationship between the Trump administration and the major technology companies, tracing the emergence of a veritable ‘algocracy’ — a system in which political power is increasingly exercised through the algorithms of digital platforms and the direct integration between Big Tech and the military apparatus, from the Anthropic-Pentagon case to nuclear crisis simulations conducted using advanced language models. The chapter concludes with an analysis of the material and environmental costs of this infrastructure — energy, water, rare earths — with a focus on the race for Arctic resources. The second chapter focuses on Palantir Technologies, tracing its history from Peter Thiel’s so-called ‘PayPal Mafia’, the evolution of its market capitalisation, and its role as an entrepreneurial paradigm in the convergence between Silicon Valley and the US national security apparatus. The third chapter examines, through a review of key cases, the practical use of Palantir’s software in Middle Eastern theatres of operations, including operations conducted against Iran, before discussing the ethical and legal implications of these technologies, with particular attention to the violation of the right to privacy. Overall, the thesis aims to demonstrate how artificial intelligence is not merely a tool at the disposal of states, but a factor that redefines the boundaries of their sovereignty, shifting decisions that directly affect the security, freedom and fundamental rights of individuals to private entities — often lacking democratic legitimacy.
Il presente lavoro indaga il rapporto tra intelligenza artificiale e privatizzazione della guerra, muovendo dall'ipotesi che l'ascesa di sistemi di IA applicati alla sicurezza nazionale non rappresenti una semplice evoluzione tecnologica, ma il sintomo di una trasformazione più profonda: il progressivo trasferimento di funzioni tradizionalmente statali — decisione strategica, sorveglianza, targeting militare — nelle mani di attori privati. Il primo capitolo costruisce la cornice teorica e storica della tesi. Dopo aver definito l'intelligenza artificiale come costruzione socio-tecnica, ripercorre il crollo del paradigma westfaliano — accelerato dagli attentati dell'11 settembre 2001 — che ha reso possibile il passaggio da una guerra fondata su eserciti e territori a una guerra combattuta nel dominio dell'informazione. Analizza quindi le “zone grigie” del diritto internazionale (legittima difesa, armi nucleari, operazioni below the threshold), che consentono di condurre operazioni ostili senza superare la soglia formale dell'attacco armato, e il rapporto tra l'amministrazione Trump e le grandi aziende tecnologiche, ricostruendo la nascita di una vera e propria algocrazia — un sistema in cui il potere politico si esercita sempre più attraverso gli algoritmi delle piattaforme digitali e l'integrazione diretta tra Big Tech e apparato militare, dal caso Anthropic-Pentagono alle simulazioni di crisi nucleare condotte su modelli linguistici avanzati. Il capitolo si chiude con un'analisi del costo materiale e ambientale di questa infrastruttura — energia, acqua, terre rare — con un focus sulla corsa alle risorse artiche. Il secondo capitolo si concentra su Palantir Technologies, ricostruendone la storia a partire dalla cosiddetta PayPal Mafia di Peter Thiel, l'evoluzione della sua capitalizzazione di mercato e il suo ruolo di paradigma imprenditoriale nella fusione tra Silicon Valley e apparato di sicurezza nazionale statunitense. Il terzo capitolo esamina, attraverso una rassegna dei principali casi, l'impiego concreto dei software di Palantir in teatri operativi mediorientali, incluse le operazioni condotte contro l'Iran, per poi discutere le implicazioni etiche e giuridiche di queste tecnologie, con particolare attenzione alla violazione del diritto alla privacy. Nel complesso, la tesi intende mostrare come l'intelligenza artificiale non sia un semplice strumento a disposizione degli Stati, ma un fattore che ne ridefinisce i confini di sovranità, spostando verso soggetti privati — spesso privi di legittimazione democratica — decisioni che incidono direttamente sulla sicurezza, la libertà e i diritti fondamentali degli individui.
Intelligenza artificiale, sicurezza e geopolitica: il ruolo di Palantir nella strategia degli Stati Uniti
TOSETTO, GIULIA
2025/2026
Abstract
This paper examines the relationship between artificial intelligence and the privatisation of war, based on the hypothesis that the rise of AI systems applied to national security does not merely represent a technological evolution, but is a symptom of a more profound transformation: the gradual transfer of functions traditionally carried out by the state — strategic decision-making, surveillance and military targeting — into the hands of private actors. The first chapter establishes the theoretical and historical framework of the thesis. After defining artificial intelligence as a socio-technical construct, it traces the collapse of the Westphalian paradigm — accelerated by the attacks of 11 September 2001 — which paved the way for the transition from warfare based on armies and territories to warfare waged in the information domain. It then analyses the ‘grey areas’ of international law (self-defence, nuclear weapons, ‘below-the-threshold’ operations), which allow hostile operations to be conducted without crossing the formal threshold of an armed attack, and the relationship between the Trump administration and the major technology companies, tracing the emergence of a veritable ‘algocracy’ — a system in which political power is increasingly exercised through the algorithms of digital platforms and the direct integration between Big Tech and the military apparatus, from the Anthropic-Pentagon case to nuclear crisis simulations conducted using advanced language models. The chapter concludes with an analysis of the material and environmental costs of this infrastructure — energy, water, rare earths — with a focus on the race for Arctic resources. The second chapter focuses on Palantir Technologies, tracing its history from Peter Thiel’s so-called ‘PayPal Mafia’, the evolution of its market capitalisation, and its role as an entrepreneurial paradigm in the convergence between Silicon Valley and the US national security apparatus. The third chapter examines, through a review of key cases, the practical use of Palantir’s software in Middle Eastern theatres of operations, including operations conducted against Iran, before discussing the ethical and legal implications of these technologies, with particular attention to the violation of the right to privacy. Overall, the thesis aims to demonstrate how artificial intelligence is not merely a tool at the disposal of states, but a factor that redefines the boundaries of their sovereignty, shifting decisions that directly affect the security, freedom and fundamental rights of individuals to private entities — often lacking democratic legitimacy.| File | Dimensione | Formato | |
|---|---|---|---|
|
Il ruolo di Palantir nella stretegia degli USA.pdf
accesso aperto
Dimensione
1.16 MB
Formato
Adobe PDF
|
1.16 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/117877