On the Case of Palantir and the AI That Oppresses and Kills
In the historical period we inhabit, a certain disillusionment is rising among many people who saw information and communication technologies – and the internet in particular – as a great opportunity for general emancipation. “Something went wrong,” we often hear repeated (Luna 2025).
To be honest, we cannot deny the great enthusiasm and hopes that have supported and still support, in a certain way, the digital transition in almost every sector of human activities. However, as media scholars and researchers of cultural processes, we have never hidden the complexity and delicacy involved in negotiating the factors necessary to balance and manage the various interests that come into play in constructing a new social reality.
Nevertheless, all of us, in daily practice, have constantly “came to terms,” consistently and almost uncritically, with technologies that, when extended and adopted on a large scale, alter the way we think, experience, and practice culture (Striphas 2023).
Whatever we may mean by culture – values, material/immaterial artifacts, or ways of life – it is undeniable today how computational processes influence the work of selection, classification, and hierarchization of ideas, things, and people, while at the same time pushing us to adapt our behaviors in accordance with the new socio-technical context.
However, to avoid falling into easy dystopianism that cannot explain technological development in the complexity and historicity of social action, it is useful to start from some considerations on that alchemical forge that allows us to design and project, in the context of an algorithmic culture, the applications that are marking our living.
Opportunities and threats in a Networked Society
At the dawn of the “network society,” sociologist Manuel Castells (2001) noted how the internet laboratory is a product that over time has seen a layering of values due to the coexistence of multiple cultures, skills, interests, and passions.
Its openness and plasticity to innovative ideas and practices, its untamable and elusive technicality, its social virality, and its entrepreneurial fertility are characteristics inherited over decades in which four different temporal phases succeeded one another – defined by Castells as the scientists’ (1950s-60s), hackers’ (1970s-80s), virtual communities’ (1980s-90s), and dot.com entrepreneurs’ (1990s-2000s).
All this to emphasize how this potential for action is value-laden in its complexity and imaginative virtuality, for developing products/projects that can be made available both to dynamics open to multiple possible worlds – the pluriverse, that is, the different possibilities or ways of life of people and communities – and to colonizing and unique life systems that, for anthropologist and technologist Arturo Escobar (2018), continue to perpetuate with other means the usual patriarchal and capitalist culture dedicated to control and domination over every other form of existence.
In any case, Castells warned us that when a society tends to function more and more in and through the network, the capacity to process information comes to assume central economic relevance. He defined this new driving economy as “informationalism” – and the market value and geopolitical weight of the major hi-tech companies established in the current world confirm his theses.
The social spread of the internet has indeed allowed the growth of a new economy. In it, various companies through a systematic work of data collection on all sorts of activities (datafication), their transformation into any commodity (commodification), and their selection (selection) for commercial launch based on economic and/or strategic criteria, have relegated us – or even imprisoned (lock-in) in most online activities – in the so-called platform society (van Dijck, Poell, de Waal 2019).
It should be emphasized that, unfortunately, these platforms are mostly owned by American companies, enterprises quick to conquer, for the “network effect,” global scale dimensions, strong with solid expertise in software and investment strategies in hi-tech. The problem of having such a geopolitical imbalance concerns the accumulation of power that can easily, on their part or on the part of the political power that controls them, degenerate into various forms of exploitation/dominion.
The hypertrophic growth and consolidation of ICT (Information and Communication Technology) infrastructure necessary to support online activities and data produced by billions of people has also enabled the development, starting from the 2010s of the 21st century, of particular Artificial Intelligence (AI) algorithms, a story we have told in another article.
With the rise of AI that self-learns (machine learning) from training examples guided by big data to shape the so-called neural networks – and thus through computational models forged on a statistical basis – algorithms capable of automating/predicting functions/processes flourish, opening in the internet a new phase that we could define, for many reasons, that of techno-authoritarianism or technocracy.
When we speak of techno-authoritarianism/technocracy we speak of the welding between the political and ideological projects of certain parties/political figures and hi-tech companies offering technologies that can be fundamentally declined in terms of control/domination to organize/manage problems involving interests of public relevance. But before deepening this aspect, let’s see in general what problems the rise of AI and its autonomous agents entails.
AI and Responsibility
As we said, today social and cultural dynamics live on human agency and algorithmic agency, something also quite evident, as philosopher Rosy Braidotti reminds us, from the fact that we are continuously asked to demonstrate, in our online wandering, that we are not a robot (2019).
In this same context, AI presents itself as a new growing resource of artificial agency (thus not requiring human intervention) that is interactive, autonomous, and often self-learning. The extension and diffusion in society of intelligent and autonomous machines, as philosopher Luciano Floridi explains, determines a divorce between artificial action and the natural intelligence proper to a human being (2021).
Previously, agere, intelligere, and intentionality – intentionality of acting (acting for a purpose), intentionality in acting (acting consciously of oneself) – were three inseparable aspects of the same phenomenon. Now accountability – giving account of the action itself as the cause of something – separates from responsibility – the moral and social duty to do or control something in all its aspects.
Put in these terms, it is clear that our system of existing as mere users of a platform society – substantially guided by processes, products, algorithms, and now also AI agents made available by various hi-tech giants – Meta/Facebook, Alphabet/Google, Apple, Amazon, Microsoft, Alibaba, OpenAI, etc. – is not socially and morally sustainable.
For example, until today these platforms – also legally defended to protect their development from the power of traditional industries – have always shielded themselves regarding what happens on the network by defining themselves as simple providers of tools, referring to users as the true responsible parties for actions/content.
But we all know by now that AI applications, like all other artifacts, are not neutral: they are trained on certain data, reflect certain cultures and design constraints, carry a complexity that escapes even their designers, and are wrapped in informational opacity due fundamentally to the marketing policies necessary to support them – constantly navigating experimentation and consumption of enormous computational resources – their high development costs.
The Hype of AI and Techno-Authoritarianism
Around AI developments, a great global battle is being played out driven by the conviction that these new algorithmic systems can produce on a large scale – inserting them into the contexts of many activities (industrial, medical, recreational, educational, etc.) – various (and presumed) advantages in terms of greater efficiency, decision-making speed, control, and dominion.
Techno-authoritarianism is the belief that they can also replace – in state and public policy spheres – those considered cumbersome and faltering deliberative procedures based on democracy.
It is no longer a taboo to publicly meet hi-tech entrepreneurs who also wear the clothes of ideologues/politicians ready, literally, to “call to arms” to make use of such advantages and anticipate the “enemy of the moment.”
The Palantir Case
Among the most cited in this regard are Peter Thiel and Alex Karp, respectively president and CEO of the American company Palantir, a company that sells various types of software platforms to private individuals and public administrations to conduct analysis – with data made available to them – also in a predictive sense regarding processes of various natures.
The two entrepreneurs are also graduates in philosophical and sociological studies from renowned universities and are authors of disturbing essays on the new technological doctrine that should guide the constitution of the democratic republics of the Western alliance, the only “region” of the world, in their opinion, that can boast the consolidation of a living that can be defined as civilized.
On Thiel and Karp, the journalism is now vast and we will not deepen it, referring to the journalistic videos in the attachment, which we consider exhaustive. However, we cannot resist noting how there is a reputational dazzlement in their regard that, if the moral point of view still counts, leaves one speechless, especially when admiration comes from the political world.
In any case, strong from their positioning in business alongside the powerful US administration on many fronts, especially those of migration policy and defense – from those parts, and unfortunately not only, now renamed “of war” – the two do not hide their controversial thought at all, also justifying – in actions where their hi-tech and AI corporate technologies offer practical support – the many collateral killings of civilians – see Karp’s statements on Gaza (Nova Lectio 2026, min. 00:52).
Or when they express the sense of their “political” operation through the force of technology. In a conference, Thiel expressed himself exactly in these terms: “the basic idea was that we could never win an election to get certain things because we would be a narrow minority, but maybe we could change the world unilaterally without always having to convince people, begging and pleading with people who will never agree with us, and we will do it with technological means, and here I think technology is an incredible alternative to politics” (Nova Lectio 2026, min. 11:14).
How Palantir and its technologies have managed to hybridize over time with the power and public administrations of some Western nations, primarily with the US administration, currently led by Trump, is truly a case study through which to reread recent events on American soil – they played a decisive role in matters such as ICE for migration policies, in the birth of the new DOGE department for dismantling national public agencies, that led by Elon Musk with the slogan of a presumed efficiency and in the new reorganization and mobilization of the US army with AI.
The Greatest heist of All Time
The enormous burst of executive orders activated immediately by the Trump administration (2025) can be read as an effective strategy to disorient the judicial power, trying, with rapid blitzes, to sift through and acquire data contained in the various and compartmentalized federal databases, data then made available to various external suppliers of software platforms – the databases are differentiated and physically separated intentionally to limit access only to public officials of the specific sector for privacy reasons towards American citizens.
In defiance of the power and procedures of law, always lagging behind what we then find already cooked by technology, Palantir – a private entity – found itself suddenly – with the excuse of DOGE’s new needs – in possession of all information on American citizens held by the federal state, sensitive and private information that could now be cross-referenced with all others extractable on the web (web browsing, social media, cell phones, biometric devices, etc.).
The appropriation of data for the construction of a single database available for virtually total control over all aspects of the lives of 300 million Americans was evidently the primary objective. The urgent needs of DOGE were only an excuse, and indeed the entity, having obtained the data exfiltration, was immediately abandoned by Musk – regarding this, in the attached video service, obtaining this data is spoken of as the greatest heist of all time.
While various judicial processes were opened on the case expressing themselves unfavorably about the Trump administration’s blitzes, these finally prevailed with a narrow victory obtained only by reaching the Supreme Court, which has a Republican majority. Meanwhile, a presidential executive order was also signed entrusting Palantir with the construction of a single federal database – and anyway the data had already all been used, for example, in systems made available by Palantir for the infamous ICE operations.
These events must therefore remind us, once again, how it is necessary to be aware of the data that concerns us and that, with our digital activities, we leave on the net knowing that there are habits and techniques that can contain their risks.
At the same time, we are warned to keep an eye on the moves that the rulers of our country take regarding the safeguarding of data that concerns us and that public administrations hold, preventing them from ending up in the hands of private companies or other states – the issue of data sovereignty has always been important but with AI it becomes even more so.
On the Responsibility for Building Socio-Technical Devices
A question has always been central in the construction of systems that foresee in their functioning automatic mechanisms or, more generally, regarding complex socio-technical devices that arrange and act on wide spatial and/or temporal gaps. As a community, we feel and also have the social and moral obligation to assign responsibility to respond appropriately in case of malfunctions, also to reduce or eliminate, as much as possible, functional risks.
In her research, scholar Madaleine Elish (2019) reflected on the concept of human-in-the-loop – that is, on the necessity of always leaving a human being within automated systems – so that one can intervene manually in case of sudden inconveniences.
To understand the logic and dynamics of implementing this strategy, the scholar deepened the case of the introduction of automatic pilot systems in aircraft in the 1970s of the 20th century. As we know by now, these devices are used a great deal on aircraft, and practically the greater part of a flight is managed, if not in the phases of takeoff and landing, by these systems.
The problem that arose in implementing automatizations was the loss of “living” experience that pilots, getting used to these systems, would suffer over time, which could put them in difficulty in case of having to suddenly take over, for any anomalies, to correctly restore operativity.
For the scholar, we can therefore keep people “in the loop” but, leaving them experientially depowered, they become only scapegoats to absorb the community’s wrath – Elish sees them in the role of “moral sponges” and speaks of them, in terms of cynical design, as a “moral crumple zone” since the “crumple zone” is the absorption zone created in vehicles to absorb any impacts in case of accidents.
Regarding this, we remember “Captain Sully” – Commander Chesley Sullenberger – famous for performing the “Miracle on the Hudson” on January 15, 2009. At the controls of US Airways Flight 1549, after takeoff from New York, the plane was hit by a flock of geese (bird strike), losing both engines. Without thrust to return to the airport, he managed to land in the Hudson River, saving all 155 people on board.
Well, very likely it was a miracle due above all to the fact that the same pilot was also a highly regarded instructor regularly engaged in teaching crisis management techniques to his own colleagues.
AI as Offset: Compensation Strategies in the Military Sphere
In one of the annual presentations of its products – AIPCon 9 – Palantir invited as a speaker one of its best clients, namely the US Department of Defense (now War), in the person of Cameron Stanley, who works there as head of AI.
In a presentation titled “Multi-Domain Artificial Intelligence: The Future of Command and Control,” Stanley speaks enthusiastically about Palantir’s MAVEN platform – we remember that the development of this platform had been entrusted by the Department of Defense to Google in 2016. Following various demonstrations and also dismissals by its own personnel, unwilling to work for military objectives, Google finally gave up the contract, then entrusted to Palantir in 2018.
In the attached presentation (March 2026), which is on Palantir’s YouTube channel, Stanley speaks in many different languages as the multi-language dubbing function made via AI, new YouTube functionality, was activated.
This is to say how much Palantir cares to spread the beauty of its technologies to the maximum audience, evidently encouraging adoption by some other military apparatus in the world – moreover, the platform is already in use by the Israeli army (IDF).
When the video speaks of offset, it speaks of the concept of compensation strategies in the military sphere, that is, of how a technological innovation can compensate for a strategic impairment. In this light, nuclear deterrence was worked on as the first compensation strategy to have, compared to the enemy, a reduced army.
Once the enemy was also equipped with the atomic weapon, the subsequent compensation was implemented by developing precise, fast, and also invisible (stealth) ballistic systems. Now, Stanley explains, the cutting-hedge of compensation is AI since it is capable of guiding the remodulation and military organization, gaining efficiency/capacity/decision-making speed.
In the video, we can thus see how the Maven platform works. This is concerned with analyzing, selecting, and preparing – in all necessary steps up to presenting the “ok” button for the decision to fire – the possible targets to hit, to close what is described, without beating around the bush, as the “kill chain.”
Maven, it is explained, brings together in a single interface, uniting them, different systems so that, with simple click of mouse movements, the various targets to hit are analyzed in every detail with assigned priorities and executive characteristics, including evaluations on the quality of information and probabilities of success in case of attack.
Stanley shows how they started from chaotic operating conditions – in this command post (Scarlett Dragon) where about 2,000 military analysts operated, set up in an army base using information and material relating to the war conducted in Iraq – to be able to process, with only 20 people, 1,000 targets to hit every hour.
It took some time, he says triumphantly, but we succeeded, especially thanks to AI, which also takes care of sifting through, freeing them from the boring work of previous analysts, the thousands of hours of images captured by satellites, aircraft, and/or drones.
And every day, he states satisfied, thanks to Palantir we manage to see continuous improvements, keeping our soldiers safe and bringing them home sound and safe.
To be clear, the system we see seems like a video game, and it is so true that now warring armies, in this new atmosphere, have fun matching systems and military actions with names in fantasy style. Unfortunately, their effects do not remain in virtual reality, but kill real people (military, people defined as terrorists but now, more and more often, simple civilians including children, women, elderly).
In recent wars – Ukraine, Gaza, Iran – the images we occasionally see in video news, with targets exploding in the air, are almost certainly executions managed by platforms in Maven style, if not Palantir’s Maven itself.
The War Incident
In this presentation that exalts care, speed, and precision in preparing attack targets, there is something ironic and tragic if we think that it happened only a few days after the war named “Epic Fury” by the United States-Israel against Iran broke out. And there, right on the first day of war, February 28, 2026, in Minab, south of Iran, the Shajareh Tayyebeh elementary school was bombed and destroyed.
The presumed victims are between 175 and 180 people, most of them girls between seven and twelve years old. Currently in the West, these children are hardly spoken of anymore, evidently children of others far away –the Iranian ministers who fly around the world following peace negotiations carry with them, on planes to make them remember, the photographic images affixed to empty chairs.
Yet in those days the clamor of the fact was enormous: how is it possible that the most professional, most technologically advanced, and most subsidized armies in the world could have committed a crime of this kind?
What is immediately apparent to the eye is that the school is located near a military barracks from which it is nonetheless separated by walls for at least 10 years. Moreover, images of the building and its transformations over the years are also documented by searching as normal users through Google Maps’ web system.
The news of this horrible attack provokes unanimous outrage throughout the world while both the United States and Israel deny any involvement. Moreover, President Trump even goes so far as to accuse the Iranians themselves alluding to their poor military precision and thus to a domestic incident of a nation accustomed, adds his minister Hegseth, to tormenting its own civilians. TV service from channel 7 (Piazza pulita, 2026, 10 minutes).
The Dispute over Responsibility for the War Crime
However, videos documenting the attack begin to circulate on social media. Being multiple, and therefore carried out in several phases, some people in the area had the opportunity to film part of the action with a cell phone, from which it is evident that a Tomahawk missile struck.
International journalism, in particular the BBC, begins to investigate and press American military high officials, the same Trump and his war minister Hegseth, in particular regarding the fact that it is an American missile. The American embarrassment is evident but Trump states that missiles of that type are also had by other countries since they sell them.
Everyone understands that anyway the Iranians could not have had them and with this the hypothesis begins to advance that something, in the American war machine, went wrong. From that moment on, to questions about the event, US authorities entrench themselves behind “there are investigations underway.”
The BBC tries to deepen the possible origin of the incident involving military experts and also international law consultants to understand how these attack operations are normally managed.
Indeed, there is unanimity of views on the fact that targeting a school is implausible given that satellite images, in this case, are easy to read since there are easily readable behavioral patterns such as the comings and goings in the school of parents who take and pick up their children, in addition to the activities of the children in the external playing fields being easily documented.
What comes out in discussions with “those in the know,” however, is also concern for the reorganization that the US army has been undergoing for some time, with the reduction of personnel, also that dedicated to controlling the rules of engagement, generally military lawyers who evaluate, in the face of a possible action, whether the consequences are in line with the laws and recommendations formulated to contain collateral damage – which means safeguarding possible civilian deaths.
And in this work of efficiency, journalists discover that the army is increasingly introducing automated functionalities with AI, a fact on which, however, there is no documentation or public debate, and on which therefore it is difficult to obtain precise documentation.
Whatever the cause that generated the incident, there seem to be no doubts that it is a war crime that must be prosecuted by law as the authors have demonstrated to have been negligent in incorrectly identifying a civilian target.
Unfortunately, it is noted, the United States, not adhering to the International Criminal Court, will not be judged and will treat the case internally through their own military courts, possibly prosecuting their own military but also based on the consideration – as is their practice – that, in case of non-intentionality – that is, of not having had knowledge that it was a civilian target – it is an involuntary offense and therefore not assimilable to a war crime.
AI Takes the Blame But…
In the same days in which the tragedy occurred, there had been on the press a dispute between the Pentagon and Anthropic, an AI company famous for developing an AI LLM (Large Language Model) model that makes its chatbot Claude talk. In order to collaborate with the military, intending to integrate the technology into their systems, the company had placed conditions – not to use them for purposes related to mass surveillance and killings.
In the AI landscape, Anthropic is famous for wanting to differentiate itself in terms of ethics with its competitors, an operation that can also be read as marketing – we all remember Google’s motto, “don’t be evil,” launched as a fundamental promise in wanting to always act in the interest of users who entrust them with much of their privacy.
The coincidence of the controversy directs the American public debate on the possible causes of the school bombing precisely towards AI – in the hype of AI now everything is AI! Perhaps it is a possible case of “hallucination” by Claude regarding the selection of a wrong target?
The search for the cause in AI in the public debate stimulates an acute journalist of the British newspaper The Guardian, Kevin T. Baker, who, almost a month later, on March 26, 2026, publishes a long investigative article to take stock of the situation.
The title of the article “AI got the blame for the Iran school bombing. The truth is far more worrying” sets the tone for an investigation that retraces the long-term steps of American political and military strategy.
Meanwhile, it is clarified how Anthropic’s AI, with its LLM that animates the chatbot Claude, was used only to allow the generation, upon request, of consultation reports relating to data associated with systems: it does not therefore enter into the merit of decisions that instead belong to the targeting system (kill chain) Maven, built by Palantir.
But for Baker in this event – in which a highly automated system with AI techniques selecting and preparing targets to offer to the trigger of a human being does not notice it is pointing at an elementary school – AI is not the most important point.
Indeed, AI is currently “a charismatic technology” that has the power to shape “the entire field around it, as a magnet organizes iron filings,” distracting us from true human responsibilities.
In fact, American history has a long tradition of discussions and also failures in developing attack systems that act from above – abstracting and recombining informational elements coming from the field – without having real feedback of what happens in a territory that reorganizes to respond appropriately.
And this is true since the Vietnam War, in which the US suffered burning “technological” defeats if compared to the “low intensity” technologies available to the enemy.
The real question that for Baker we should ask ourselves is instead relative to what happened to the organization – to the bureaucratic framework to organize the passages between detecting something and destroying it – that is, “to the kill chain, and the answer is Palantir.”
In general, the military have worked for at least 80 years to make this chain as short as possible, and Palantir’s intelligent Maven system is the latest iteration of this compression.
The Maven project was born when generals realized that 80% of military analysts’ time was absorbed by tasks considered trivial, such as supervising films and images of territories to attack that lasted hundreds of hours. The fundamental premise of the project was that “the machine could watch so that the analyst could think.”
So here we think of organizing a military AI department. The XVIII Airborne Corps began testing the system in an exercise called Scarlet Dragon, which started in 2020 as a war tabletop exercise in a windowless basement at Fort Bragg.
The objective was to test whether the system could give a small team the targeting capacity that previously required thousands of people.
Over the next five years, Scarlet Dragon grew into a military exercise using live ammunition, spanning multiple states and branches of the armed forces, with “forward-deployed engineers” from Palantir and other contractors embedded alongside soldiers. Each time the exercise was run, it was meant to answer the same question: how fast could the system move from detection to decision? The benchmark was the 2003 invasion of Iraq, where roughly 2,000 people worked the targeting process for the entire war. During Scarlet Dragon, 20 soldiers using Maven handled the same volume of work. By 2024, the stated goal was 1,000 targeting decisions in an hour. That is 3.6 seconds per decision, or from the individual “targeteer’s” perspective, one decision every 72 seconds.
Before Maven, operators worked on eight or nine separate systems simultaneously, extracting data from one, cross-referencing data in another, manually moving detections between platforms to assemble the intelligence and approvals necessary for each attack.
Maven has consolidated all this behind a single interface that filters and decides on inputs from different systems, converting a data point on the map into a formal detection to move it into a targeting pipeline.
These targets then move through columns representing different decision-making processes and rules of engagement. The system recommends how to hit each target – which aircraft, drone, or missile to use, which weapon to associate with it – what the military call a “course of action.”
The officer selects among the classified options, and the system, depending on who uses it, sends the target package to an officer for approval or moves it to execution.
In organizations that function on formal procedures, we instead need someone within the process to interpret the rules, notice exceptions, recognize when categories no longer fit the case. In short, Baker notes, windows are created where errors can be caught.
But if we place ourselves within the framework of efficiency, all this is recorded as latency and therefore defect. As Baker states, Scarlet Dragon in fact succeeded in eliminating discussions on targeting, deliberation processes, hesitation, and moments in which someone had time to object or notice that something was wrong.
In this way, even the tremendous responsibility of deciding on life and death is increasingly delegated – expelling as many people as possible from the loop – to the mathematical rigidity of procedures codified in the opacity of software – creating an “Artificial Bureaucracy” – that leaves the few human operators in the role of probable, as Elish says, “moral sponges.”
But AI has also served to distract, Baker concludes, from the real issues at stake – cynically shielding the community from its multiple co-responsibilities.
Congress did not authorise this war. In two weeks, American forces struck 6,000 targets. The school was one of them. American forces killed almost 200 people, and the reporting reached for “AI error”, which domesticated the event into something a better algorithm or better guardrails could have prevented.
In the days after the strike, the charisma of AI organised the entire political conversation around the technology: whether Claude hallucinated, whether the model was aligned, whether Anthropic bore responsibility for its deployment. The constitutional question of who authorised this war and the legal question of whether this strike constitutes a war crime were displaced by a technical question that is easier to ask and impossible to answer in the terms it set. The Claude debate absorbed the energy. That is what charisma does.
It has also occluded something deeper: the human decisions that led to the killing of between 175 and 180 people, most of them girls between the ages of seven and 12. Someone decided to compress the kill chain. Someone decided that deliberation was latency. Someone decided to build a system that produces 1,000 targeting decisions an hour and call them high-quality. Someone decided to start this war. Several hundred people are sitting on Capitol Hill, refusing to stop it. Calling it an “AI problem” gives those decisions, and those people, a place to hide.
References
Baker, K. T., 2026, “AI got the blame for the Iran school bombing. The truth is far more worrying“, 26/03, theguardian.com.
Braidotti, R., 2019, Posthuman Knowledge, London, Polity Press.
BBC News, 2026, “Was the strike on a primary school in Iran a war crime? | The Global Story“, marzo, youtube.com.
Castells. M., 2001, Internet Galaxy, Oxford, Oxford University Press.
Elish, M. C., 2019, “Moral Crumple Zones: Cautionary Tales in Human Robot Interaction“, in Engaging Science, Technology, and Society, v. 5, pp. 40-60.
Epoch Philosophy, 2026, “Palantir and the Conspiracy to Own Everything: The Largest Heist in History“, youtube.com.
Escobar, A., 2018, Designs for the Pluriverse: Radical Interdependence, Autonomy, and the Making of Worlds, New York, Duke Univ Press.
Floridi, L., Cabitza, F., 2021, Intelligenza artificiale. L’uso delle nuove macchine, Milano, Bompiani.
LA7 Attualità, 2026, “’Ho trovato mio figlio sotto le macerie’. L’attacco alla scuola iraniana di Minab“, youtube.com.
Luna, R., 2025, Qualcosa è andato storto. Come i social network e l’intelligenza artificiale ci hanno rubato il futuro, Milano, Solferino.
Nova Lectio, 2026, “L’azienda che spia l’Occidente e decide le guerre“, youtube.com.
Stanley, C., 2026, “Intelligenza artificiale multidominio: il futuro del comando e del controllo | CDAO all’AIPCon 9“, youtube.com.
Striphas, T., 2023, Algorithmic Culture Before The Internet, New York, Columbia University Press.
Van Dijck, J., Poell, T., De Waal, M., 2018, The Platform Society: Public Values in a Connective World, Oxford, Oxford University Press.
