Writing about power or change in a world that moves so fast that energy and mass struggle to keep up has become nearly impossible. Every day now, we wake up to geopolitical shifts that leave us bewildered because they seem too sudden.
But neither the speed of current events nor the diplomatic lability of certain key players in international affairs need distract us from clear trends emerging in 21st-century governance. They centre around two related matters: autocracy and technology, with a marked emphasis on artificial intelligence (AI). A number of influential authors, journalists, and researchers insist that, aside from climate change, these perils are perhaps irrevocably altering our world, along with the post-war realities with which we were familiar.
Russia: An Old Template for a New Century?
Vladimir Putin’s obsession with absolute power provides a direct entry into an examination of authoritarianism and technology in Giuliano da Empoli’s novel, The Wizard of the Kremlin. Da Empoli, a Swiss and Italian professor, essayist, novelist, and political advisor to Italian ministers, has been hailed as the writer world leaders love to read, not so much because they see him as a modern-day Machiavelli, but because he holds a puzzled mirror to their lives. A glacial reflection of Putinism, The Wizard of the Kremlin acts as a Trojan horse, ushering Western readers into the wings of privilege in Moscow and Saint Petersburg while toying with truth and fiction.
Da Empoli’s protagonist, Vadim Baranov—a cold fictional spin doctor to a semi-fictional Putin, also known as “the Tsar”—warns us about the ultimate dictator of the future. “Human history ends with us,” he announces,
with you, with me, and maybe with your children. Afterwards, there will be something, but it won’t be humanity. The beings that come after us, if there are any, will have different ideas and preoccupations than those that have engaged man until now.
Baranov’s vision is that robots, and AI within them, will transform and discipline humankind, eliminating any deviation from the norm, and leaving sovereign power to prevail without effort. His dystopia is religious in nature, introducing AI and robots as incarnations of God, finally descended from Heaven to meet His creatures and impose His version of peace upon their turbulent lives.
The idea is Baranov’s, not Putin’s, and it is far from new, but this iconoclastic novel, which won the Grand Prix du Roman de l’Académie Française and was a finalist for the Goncourt Prize in 2022, also demands that we consider what makes Vladimir Putin a particularly successful despot: a capacity to push ruthlessness and divisive chaos to a critical point so that abject terror prevails domestically and abroad.
Da Empoli’s understanding of unlimited power and the relationship between autocracy and AI is further developed in his new non-fictional volume, The Hour of the Predator: Encounters with the Autocrats and Tech Billionaires Taking Over the World.
Next to the “predators” da Empoli introduces in this last work, Putin himself seems anachronistic, an aging emperor who has lost touch with the times and fights his wars the old-fashioned way, through ruse, boots on the ground, and hesitant use of technology.
The New Kings and Their Brand-New God: AI
The Hour of the Predator begins where The Wizard of the Kremlin ends and asks whether we are ready to let go of democracy and be ruled by autocrats and, next, by machines.
Having dealt with a stylized Putin in his fictional work, da Empoli turns his attention to contemporary and particularly histrionic heads of state, wannabe monarchs who feel equally entitled to supreme power, namely the notorious Donald Trump, whom he calls a “functional illiterate”; the chillingly psychopathic Mohammed bin Salman, crown prince of Saudi Arabia, described as a “friendly giant,” all “sweetness and light” as he tortures enemies; and Nayib Bukele Ortez, the president of El Salvador, who suspended legal and procedural rights and incarcerated over 85,000 tattooed persons whom he accused of being pandilleros, or gang members, even if many were not. For such heads of state, as for Putin, the medium—cruelty—is indeed the message.
Da Empoli describes these leaders as “Borgians,” or honorary descendants of the corrupt and murderous Borgia family during the Renaissance. Chaos to rule is also their favourite game. They ensure the loyalty of a critical segment of their people through fear, by running roughshod over civil liberties, and by invoking traditionalist necessities: inflation control, fighting crime, curbing immigration, and helping themselves to the resources of enemy nations.
Chaos to rule is also their favourite game.
Where da Empoli shines is in his opening as he observes that what distinguishes these new autocrats from Putin is that they are patrons of technology, admirers of AI.
He compares them to the Aztec ruler Moctezuma II who, upon meeting the Spaniards led by Hernán Cortez in 1519, was so mesmerized by their armour and military technology—those “special blowpipes that spat fire and thunder”—that he mistook them for gods and allowed himself and his people to be conquered. Today, the technology happens to be AI, but little else has changed.
The “rituals of degradation” to which the Aztecs assented remain as pervasive as they are enduring. Today, so-called “tech bros” descend like deities from their planes, and politicians beg them to build “a research centre or an AI laboratory” on their territory, at times in exchange for technology deregulation.
Among current tech billionaires, “predators” in their own right, da Empoli names Elon Musk and Sam Altman, the former CEO of OpenAI, but also older magnates like Bill Gates and Eric Schmidt, former CEOs of Microsoft and Google, respectively. As tycoons of technology, they are not ushering in a hip, transient fad so much as a new age that is here to stay, the measure of which we have yet to take, and from which we might well emerge as a radically different species, as Baranov suggests at the end of The Wizard of the Kremlin.
Da Empoli’s critique of AI, which he renames “authoritarian intelligence,” is withering. He recounts a Lisbon event in May 2023 during which Sam Altman and British AI entrepreneur Demis Hassabis outlined their vision of the future to a high-IQ élite that included centenarian Henry Kissinger. “[T]he more that these tech overlords attempted to put them at their ease,” da Empoli writes, “the more the audience members could feel an icy hand caressing their spine,” for they suddenly realized that all that “constituted the essence of the human adventure—the autonomy of the individual, for starters—” was about to dissolve. Notably, political leaders began to suspect that they might be outpaced by the very technology they worship and with which they hoped to bolster their authority.
Who’s Afraid of AI?
Unfortunately, da Empoli never clarifies how we might be subjugated by robots, and this is where Madhumita Murgia’s Code Dependent: Living in the Shadow of AI helps us further understand the toxic, but occasionally redeeming, power of AI.
Born in Bangalore and raised in Mumbai, Murgia is an Indian-British journalist and author, and the first AI Editor at the Financial Times. In 2024, Code Dependent was shortlisted for the Women’s Non-Fiction Prize.
Murgia aptly re-centres AI within a human framework: it is robotic and dependent on human input; because it has been used and misused by humans, it may still be curtailed or defeated.
Her assessment of AI is as devastating as da Empoli’s. She defends AI in the realm of healthcare since it can potentially save lives, especially in remote rural areas and developing countries where access to medical technology is scarce. AI, for example, can read an X-ray like a competent radiologist, giving a general practitioner the green light to begin treatment immediately instead of waiting weeks for a medical report.
Although Murgia does not mention them, in our war-ridden era, AI-powered mine clearance systems spare human and canine lives and must be added as examples of positive technology. Murgia’s assessment of AI, however, is generally disparaging in areas that fail to benefit humanity directly.
She is especially scornful of instruments like ChatGPT. American science fiction writer Ted Chiang, whom she quotes extensively on this subject, points out that Large Language Models (LLMs) like GPT, among others, are mostly useful for repetitive and unintelligent tasks “no one necessarily wants to read or write,” which is “a statement about how much bullshit we are required to generate and deal with in our daily lives.”
We are at a crossroads where general human intelligence must understand how specialized AI works in order to defeat it when it matters.
As an aside, Chiang’s words are worth pondering in the academic sphere. Higher education worries that ChatGPT undermines academic integrity and the capacity for critical thinking. In fact, students use it to cheat on roughly the same homework that has been assigned for decades, which makes data accumulation and pattern recognition unavoidable. AI detects and reproduces such patterns faster than most humans. We are therefore at a crossroads where general human intelligence must understand how specialized AI works in order to defeat it when it matters.
Where Murgia trounces AI is in areas such as law, human rights, and civil liberties. AI exploits workers in places like Bangladesh and the Philippines, where they are paid a pittance to “train software”; it eliminates paying jobs by replacing skilled workers under the pretense of efficiency; it is a notorious polluter; it infringes authorial copyright; and it oppresses individuals by allowing for the creation of pornographic (and, most likely, political) deepfakes based on random Internet photos, robbing victims of their agency and identity. It is also capable of surprisingly accurate facial recognition that assists local policing and repressive political undertakings.
In these last instances, the potential for abuse of power in the hands of a few is particularly shocking; soon, truth, privacy, and personhood may become luxuries of the past, and the gap between the powerful and the powerless, the autocrats and the rest of us, will become unbridgeable.
War: What Is It Good For?
Murgia fails to dwell on such details, but on an apocalyptic scale, the use of AI for destructive ends that risk violating human rights and civil liberties is especially unsettling. US Secretary of War Pete Hegseth’s decision to scapegoat AI company Anthropic because it had ethical qualms about working with the Pentagon in enabling usage of autonomous weapons and mass surveillance points to the technology’s dangers in the hands of any state that can afford it.
AI has in fact been used by the US during Operation Epic Fury, the 2026 conflict between the US and Iran. The Brennan Centre for Justice reports that America has used an AI system known as Maven “to enhance intelligence analysis, surveillance, and targeting.” Maven notably runs on software conceived by Republican kingmaker and tech baron Peter Thiel’s Palantir Technologies. The AI system, however, is less than polished and is prone to “drift” as accuracy lessens under real-world conditions like sandstorms.
Instruments of war, in other words, are being used when they are still at a relatively experimental stage of their development. The battleground becomes a testing ground, and after the equally experimental “Little Boy” and “Fat Man” bombs of 6 and 9 August 1945, history’s lessons sashay past us, unheeded.
In the Middle East, AI has had lethal consequences for the last few years.
In the Middle East, AI has had lethal consequences for the last few years. Never mind that all parties to local conflicts routinely use AI for disinformation purposes. Quite beyond this, +972, Independent Journalism from Israel-Palestine reveals that Israel’s use of AI in military operations has changed how war is waged.
In a 2024 report, +972 raises the alarm about high-tech targeting systems like Lavender, developed by the Israeli Defence Forces’ Unit 8200, an elite intelligence division, and used after the Hamas attacks of 7 October 2023.
Lavender and a parallel surveillance system allegedly “clocked as many as 37,000 Palestinians as suspected militants—and their homes—for possible air strikes.” Even though intelligence services sought male militants, the bombs would tend to strike their homes at night, when they were almost certain to be there, killing other members of their families in the process.
Sam Mednick and others at Associated Press warn that, additionally, “US tech giants have quietly empowered Israel to attack and kill many more alleged militants more quickly in Gaza and Lebanon through a sharp spike in artificial intelligence and computing services.”
The US and Israel are recent cases of what looks like less-than-cautious use of contemporary technology. Soon, however, the rest of the world will follow, as the race for new systems of war mimics the old race to fathom the atom’s nucleus. Predictably, a number of innocents will die. Predictably, tech lords and their patrons will grow wealthier.
Questions these systems raise, of course, are whether they are accurate or precise, and whether they are legal. Surprisingly, few critics review the ethics and legality of technology that is used as a weapon.
Surprisingly, few critics review the ethics and legality of technology that is used as a weapon.
Both the Maven and Lavender systems, for instance, potentially violate international privacy laws by intercepting foreign communications without consent, breaching privacy rights and data protection laws.
And when a party to a conflict possesses advanced technology that the other side lacks, that party may use it disproportionately, which may give rise to human rights violations.
Finally, human commanders may find themselves overwhelmed by the speed and intensity of battle and be tempted to trust AI rather than their own judgment. This last scenario opens the door to a variation on the well-established Arendtian view of evil. The infamous exculpatory statements at Nuremberg—“I was merely obeying orders”—risk one day morphing into a modern restyling: “I trusted the smart robot’s lead.” Without guardrails, more major powers become open to this possibility.
Yet that would still be better than claiming that the robot did everything on its own, which would be the case for human-out-of-the-loop AI structures like Lethal Autonomous Weapon Systems (LAWS). The main legal challenge posed by LAWS is that when their deployment results in what amounts to a war crime, the attribution of direct or causal responsibility becomes obscured. In part for these reasons, fully autonomous lethal systems remain largely undeployed, even as a number of states are actively building them.
In this respect, Hannah Arendt’s banality of evil could turn into an asepsis of evil, where ethically debatable human reactions are replaced by AI systems whose procedural operations are designed to be incapable of guilt, moral hesitation, or reflective interruption.
We, the Majority
An additional concern with the rise of these technologies of war lies in their heavy reliance on investment. As Atlantic writers Matteo Wong and Charlie Warzel note, in its “quest for unbridled growth,” the AI industry has attracted an enormous influx of capital from investors hoping to strike it rich, which has led to the swift and premature development of technology at the expense of economic stability. Right now, Wong and Warzel argue, “all economic growth in the United States comes from AI investments.” But ongoing conflicts, and notably the blockade of the Strait of Hormuz, have already caused shortages of materials critical for AI hardware production. With supply disruptions, the AI industry will face radical strain, raising the risk of broad economic insecurity in the US and possibly worldwide.
Beneath this speculative frenzy lies a broader social reality: while AI development is driven by power, middle-class investors are in the same game, expecting shared prosperity.
In Code Dependent, Madhumita Murgia proposes a different outcome: we have developed AI and may soon wish to rein it in for the common good. Now is the time to react, though, because the technology, which is constantly being developed and perfected, is fast outrunning our capacity to understand it.
Where da Empoli damns current autocratic leadership, Murgia provides a faint glimmer of hope. She ends one of her chapters with the words of a Chinese Tiananmen Square survivor who looks back on his experience as an activist when he was told he was merely part of a minority of malcontents: “we were never the minority,” he muses. “We are the majority.” The goal of abusive power is to deny democracy and the rational observation that the people have decision-making power as a collective. But we are, objectively, statistically, the majority.
Saint Augustine reminds us that Hope has two daughters: Anger and Courage. Whether we choose to exercise our anger and our courage to curtail the combined ravages of autocracy and harmful AI depends on us and a handful of morally responsible leaders, but when that day comes, let’s hope that too much time has not gone by. The distressing reality is that we had also better hope that Socrates was dead wrong about democracy, and that a good number of citizens are informed and engaged enough to care before Vadim Baranov’s prophetic vision in The Wizard of the Kremlin comes true, and we lose our very humanity and the imperfect world we nonetheless cherish.
Bibliography
Da Empoli, Giuliano. The Hour of the Predator: Encounters with the Autocrats and Tech Billionaires Taking Over the World. Translated by Sam Taylor, Pushkin, 2025.
—. The Wizard of the Kremlin. Translated by Willard Wood, Pushkin, 2023.
European Digital Media Observatory (EDMO). “The First AI War: How the Iran-Israel Conflict Became a Battlefield for Generative Misinformation.” 14 July 2025, https://edmo.eu/publications/the-first-ai-war-how-the-iran-israel-conflict-became-a-battlefield-for-generative-misinformation/
Mednick, Sam, et al. “How US Tech Giants Supplied Israel with AI Models, Raising Questions about Tech’s Role in Warfare.” AP, 18 February, 2025, https://apnews.com/article/israel-palestinians-ai-weapons-430f6f15aab420806163558732726ad9.
Murgia, Madhumita. Code Dependent: Living in the Shadow of AI. Henry Holt, 2024.
Toh, Amos, and Emile Ayoub. “The Military’s Use of AI, Explained.” The Brennan Center for Justice, 12 March 2026, https://www.brennancenter.org/our-work/research-reports/militarys-use-ai-explained.
Wong, Matteo, and Charlie Warzel. “Welcome to a Multidimensional Economic Disaster.” The Atlantic. 26 March 2026, https://www.theatlantic.com/technology/2026/03/ai-boom-polycrisis/686559/
Yuval, Abraham. “‘Lavender’: The AI Machine Directing Israel’s Bombing Spree in Gaza. +972, Independent Journalism from Israel-Palestine. 3 April 2024, https://www.972mag.com/lavender-ai-israeli-army-gaza/
