Pope Leo XIV’s new encyclical, blessed by an AI chief, claims to govern the algorithm. But is the Vatican consecrating tech power—or challenging it? Join us on Telegram , Twitter , and VK . Contact us: info@strategic-culture.su The encyclical and its moment On May 25, 2026, in the solemnity of the Synod Hall at the Vatican, an event took place that history will remember as a watershed moment: Pope Leo XIV personally presented the first encyclical of his pontificate, Magnifica Humanitas . On the protection of the human person in the age of artificial intelligence. No pontiff had ever personally presided over the public presentation of his own doctrinal document. The gesture, in its extraordinary break with centuries-old protocol, was not accidental. It was a declaration, one of those that marks an epochal transition—not from a purely ecclesiological perspective, but rather because of the significance of the content of what was presented.
To make the event even more significant, seated beside the pope was not a cardinal of the Curia, a theologian from the Pontifical Academy of Sciences, or a philosopher from the Roman phenomenological school, but Christopher Olah, co-founder of Anthropic, the California-based company that developed the Claude model. For the first time in contemporary history, artificial intelligence has become the central and foundational theme of a new pope’s first major doctrinal document. The algorithm has received the honor previously reserved for the family, peace, and social justice.
The date affixed to the encyclical’s signature—May 15—was no less eloquent. One hundred thirty-five years earlier, on the same day in 1891, Leo XIII had promulgated Rerum Novarum, the founding document of modern Catholic social doctrine, the one that for the first time committed the Church to taking a stand against the devastation of the Industrial Revolution, child labor, the exploitation of the working masses, and the issues of private property and just wages. The current pontiff, who chose the same name as his late-19th-century predecessor with a deliberateness that excludes any symbolic ambiguity, seeks to establish an explicit and imperative equivalence: artificial intelligence is to our time what the steam engine and the factory were to the time of Marx and the first trade unions. It is the social issue of the century.
The day after the signing, on May 16, Leo XIV approved the creation of a permanent Vatican commission on artificial intelligence: for the first time in its two-thousand-year history, the Holy See is institutionalizing its relationship with AI under a single governing body. The message to the world was clear and devoid of any diplomatic ambiguity: the Church of Rome is positioning itself to be the global conscience of the algorithm. But the choice of the tech representative called upon to share the stage with the pontiff was not a ceremonial gesture. It was a statement of alignment. A new axis between the Vatican and Silicon Valley? To understand what Dario Amodei was doing in Rome in the days following the encyclical’s presentation—and why the Vatican had chosen Anthropic specifically among all the major global artificial intelligence firms—we must reconstruct the geopolitical context that precipitated this meeting with the speed of a diplomatic crisis.
On February 27, 2026, the Trump administration had signed an executive order requiring all U.S. federal agencies to immediately cease all commercial activity with Anthropic. In the hours that followed, Defense Secretary Pete Hegseth had labeled the company “a risk to the national security supply chain”—a designation never before applied, in recent American history, to a domestic private enterprise. OpenAI, Sam Altman’s company, had filled the void with surgical precision, signing a contract with the Pentagon at the very moment Anthropic was blacklisted. The dispute then moved to the courts, with conflicting rulings across various levels of jurisdiction and a case still pending.
The Vatican therefore did not choose OpenAI, even though it is the most commercially powerful brand in the sector. It did not choose Palantir, despite the fact that Peter Thiel had visited Rome the previous March for a series of closed-door seminars on the relationship between technology and democracy, received in curial circles with what observers present described as glacial coldness. Instead, it chose the only company among the major AI firms that had paid the price of exclusion from the Pentagon for refusing to remove the ethical constraints embedded in its models. The Holy See, in other words, chose the partner that the Trump White House had just rejected. And it did so on the anniversary of *Rerum Novarum*, lending the initiative the solemnity of the highest historical precedent in Catholic social teaching.
At the event in the Synod Hall, Olah had publicly expressed what is rarely heard from those who develop artificial intelligence systems: the questions raised by AI, he said, “are bigger than the research community” and cannot be left solely in the hands of scientists or companies. He had listed three urgent issues of historic significance: the risk of massive job losses, the deeply unequal distribution of economic benefits between rich and poor countries, and the growing opacity of algorithmic systems—increasingly complex models that no one, not even their creators, is able to fully understand from the inside. He added that there is an even more serious problem: the absence of any mechanism capable of equitably distributing the economic benefits of AI. “It is an unresolved problem,” he acknowledged, “and it is precisely the kind of problem that historically the Church has refused to allow the world to ignore.” A radical public self-criticism, uttered by a company worth three hundred and eighty billion dollars. Anthropic: the “ethical” company and its contradictions Founded in 2021 by Dario Amodei and his sister Daniela, along with a group of researchers who left OpenAI, Anthropic has built a public narrative in just a few years based on three pillars: safety, ethical alignment, and algorithmic transparency. Co-founder Olah embodies this image better than anyone else: a researcher who studies what happens inside neural networks, who is concerned that AI systems be understandable and governable, who advocates for the need for external oversight—by governments, religious institutions, and civil society—over technologies that no single company can responsibly manage on its own.
It is this image that made Anthropic appealing to the Vatican. And it is this image that, upon closer inspection, reveals its internal fractures. Because the very same week that Olah ascended to the Chair of Peter to receive what some commentators did not hesitate to call “the Vatican’s anointing of humanistic artificial intelligence,” accounts emerged of what had transpired in the preceding months in the other seat of power—that of the Pentagon.
According to analyses published by La Fionda and reconstructed using open-source information, the first friction between Anthropic and the U.S. military apparatus emerged in January 2026, during the operation to capture Venezuelan President Nicolás Maduro in Caracas, conducted by the CIA with the support of the Maven system—the artificial intelligence program applied to military operations—and thus, according to reports, with the assistance of the Claude model. Anthropic had formally protested, arguing that the operation exceeded the contractually agreed-upon limits of use. From that moment on, the relationship had soured. The Pentagon demanded the removal of the restrictive clauses. Anthropic resisted. The crisis had escalated the following February with Trump’s executive order.
But the same source reports that in the first twenty-four hours of the joint U.S.-Israeli airstrikes on Iran—at the very same time the Pentagon was officially removing Anthropic from its supply chain—the Claude model had helped select a thousand targets. The two news items coexist in the same week and shed a disorienting light on one another. This is not merely an irresolvable contradiction on a logical level: it is a contradiction that reveals the deep structure of the entire system in which Anthropic operates—whether it wants to or not.
The tension between Anthropic’s commercial weight and the words spoken at the Vatican was hard to ignore, and Olah did not try to hide it.
What emerges is not necessarily proof of strategic bad faith, but something more disturbing: the demonstration that the categories of “ethical company” and “military company” are by no means impermeable to one another in the contemporary American technological ecosystem. They coexist, contain one another, and contradict one another. And it is precisely this coexistence that is the structural reality no doctrinal document, however solemn, can dissolve with a symbolic anointing. The three chambers of power As the brilliant Italian journalist Margherita Furlan has observed in her recent articles, there are “three chambers.” The first is Anthropic’s meeting room in San Francisco, where algorithmic models are produced. The second is the operations room of the Maven system at the Pentagon, where these models are integrated into military decision-making chains. The third is the audience hall of Leo XIV in the Vatican, where the oldest and most authoritative symbolic institutions of the Western world confer moral legitimacy on the entire system.
The metaphor is effective not because it posits a conspiracy—it does not—but because it describes a functional structure. It is not necessary for the three rooms to communicate directly, for their occupants to agree in advance, or for secret meetings or pacts signed in the shadows to exist. Structural power, as Susan Strange had intuited in her analysis of the decline of the state in advanced economies, does not operate through explicit agreements but through convergences of interest that consolidate over time until they become the invisible grammar of the world order.
In this grammar, the role of symbolic institutions—those that hold the monopoly on moral legitimacy—has always been essential. The Catholic Church has performed this function for centuries in relation to the temporal power of kings, emperors, and great merchant families. The Peace of Westphalia, the role of the Curia in pre-industrial European diplomacy, the Holy See’s position in the conflicts of the twentieth century: all attest that the Vatican’s function as a “moral clearinghouse” has weathered political revolutions, world wars, and ideological collapses unscathed.
Today that function is readapting to platform capitalism. The question that deserves to be asked without euphemisms is whether the Vatican, in undertaking this operation, is exercising a critical and corrective power—as its namesake predecessor Leo XIII exercised it toward industrial employers with *Rerum Novarum*—or whether it is performing a legitimizing function that consolidates, rather than challenges, the system it claims to want to govern. The theological revolution of the algorithm The content of the encyclical Magnifica Humanitas deserves a reading that goes beyond a judgment on the merits of its individual positions—whether agreeable or debatable—to grasp the transformation the document introduces into the conceptual architecture of Catholic doctrine.
The title itself is revealing. “Magnifica Humanitas”—magnificent humanity—is a formula that recalls the tradition of Christian humanism, the centrality of the person created in the image and likeness of God, the inalienable dignity of the human being as the foundation of all social ethics; but when applied to artificial intelligence, that formula serves a different function: it does not defend the human person against the machine, but rather seeks to integrate the machine into the horizon of the person. It is not a critique of technology; it is an attempt to tame it theologically.
Fundamental categories of the Christian tradition are in fact reinterpreted in the document through a technological lens. Discernment—which in the Ignatian tradition is the spiritual process of distinguishing between good and bad movements in the believer’s soul—becomes a category applicable to algorithmic systems: to discern, in the new lexicon, also means evaluating the impact of technologies on human life. Conscience—which in Catholic moral theology is the inner sanctuary where the person responds directly to God—is extended to include the responsibility of organizations that develop AI. Truth—which in the scholastic tradition is the adaequatio rei et intellectus, the adequacy of the intellect to the thing—must contend with systems that produce probabilistic outputs and can generate what technicians call hallucinations.
What is emerging is not merely a lexical update or a doctrinal marketing ploy. It is something deeper and less reversible: the progressive transmigration of theological language into the orbit of techno-managerial language. Once the Church has agreed to speak of “ethical algorithms,” “model alignment,” and “AI governance” as spiritual categories, the direction of the conceptual borrowing inevitably tends to reverse. It is no longer just the Church lending its moral vocabulary to technology: it is technology that begins to lend its functional vocabulary to the Church. And when the language of salvation gives way to the language of optimization, prediction, and algorithmic management of reality, something essential has already been transformed.
There is also an issue that none of the enthusiastic commentaries on the encyclical has so far addressed with the necessary frankness: that of epistemic authority. Who, in the age of the algorithm, holds the power to determine what is true? Catholic tradition has answered this question precisely for centuries: the Magisterium of the Church, through its interpretation of Revelation, is the normative reference point for the believer’s conscience. But large language models—trained on billions of texts, capable of producing plausible answers on any topic, accessible to anyone with a smartphone—are becoming, in the daily lives of hundreds of millions of people, a new form of epistemic authority. Undeclared, unconsecrated, accountable to no institution. Yet effectively operative. From Rerum Novarum to Magnifica Humanitas The parallel between Leo XIII and Leo XIV, between the 1891 Rerum Novarum and the 2026 Magnifica Humanitas, is not merely a rhetorical device. It is an interpretive key that illuminates both the similarities and, above all, the structural differences between the two historical moments.
Leo XIII wrote Rerum Novarum in a context in which the Church was clearly alien to the dominant economic power. Late-nineteenth-century industry was governed by capitalists who did not need the papal blessing to assert their legitimacy: they had built it through the market, force, and a liberal ideology that religion had largely marginalized as a relic of the past. In that context, the Church’s stance in favor of fair wages and workers’ rights was an act that ran counter to the interests of the dominant power. It came at a cost. It had real autonomy.
The current context is profoundly different. Anthropic is not an 18th-century master exploiting children in mines. It is a company worth three hundred and eighty billion dollars, whose shareholders include Amazon, Google, Sequoia Capital, BlackRock, and the Qatar Investment Authority; it already presents itself with an elaborate ethical narrative and comes to Rome not as an inconvenient interlocutor, but as a desired ally. The Vatican does not oppose this power; it seeks to negotiate with it for a position of influence within a system it does not question.
The question that the Church’s social doctrine should ask itself—and which the encyclical touches upon without answering—is structural: is it possible to govern ethically a system whose underlying economic architecture produces radical inequalities, monopolistic concentration of knowledge, and an intrinsic tendency toward military use, simply by negotiating with its more moderate players? Or is it necessary to question the system itself, its conditions of production, its governance, and its private appropriation of collective benefits?
Antonio Gramsci, in his Prison Notebooks, wrote that every hegemony is first built on the cultural plane and only later translates into political domination. The encyclical Magnifica Humanitas is exactly this: an act of cultural hegemony, an attempt to write the moral framework within which the next technological wave must operate. But an act of cultural hegemony can also be, paradoxically, an instrument of incorporation: it legitimizes its interlocutors while claiming to govern them. The risk of technocratic r eligion There is one final issue that this development forcefully raises and that no institutional celebration can neutralize: that of the progressive convergence between spiritual power and technological power, and the risk that such convergence will produce not a control of technology by ethics, but a sacralization of technology through ethics.
Transhumanist and posthumanist narratives—those that promise the overcoming of humanity’s biological limits, digital immortality, and the fusion of human and artificial intelligence—come into profound tension with the Christian tradition on every level: anthropological, eschatological, and sacramental. A human being who can be indefinitely improved, enhanced, and preserved through technology no longer needs redemption, grace, or resurrection. Death itself—the cornerstone of Christian soteriology—becomes a technical problem awaiting an engineering solution.
Yet the digital elites promoting these visions—with their peculiar blend of secular millenarianism, technological utopianism, and existential risk anxiety—are progressively occupying the symbolic space that once belonged to the great religious narratives. They speak of existential threats to humanity, of salvation through the alignment of AI, of a future in which technology will determine the survival or extinction of the species. In other words, they have adopted the formal structure of eschatological thought without its theological substance: the end of the world without the God who governs it, salvation without grace, original sin without forgiveness.
In this scenario, the risk the Vatican faces is not so much that of being deceived by Anthropic, but rather that of lending itself, unwittingly or deliberately, to a process of sacralizing techno-capitalism that uses the Church’s moral language to confer a veneer of depth on what is in reality a pure exercise of economic and strategic power. This is not a matter of assuming bad faith: it is a matter of recognizing the power of structures, which act independently of the intentions of individual actors.
The philosopher of technology Jacques Ellul warned decades ago that the supreme risk of technological civilization is not the machine rebelling against man, but the machine that man ends up worshiping—transforming efficiency into the ultimate value, optimization into a virtue, and prediction into prophecy. When the institutions that have historically safeguarded the sense of limit, finitude, and transcendence place themselves at the service of this new liturgy, it is not certain that they become its conscious officiants, but they become part of it nonetheless. Who controls meaning? The real stakes in the encounter between Leo XIV’s Vatican and Anthropic’s artificial intelligence are not of a technological nature. It does not concern the security of algorithms, nor the distribution of economic benefits, nor the restrictions on use in military contracts—though all these issues are of enormous practical importance. The real stakes are symbolic and political in the highest sense of the term: who controls the moral meaning of the ongoing technological revolution?
The scene on May 25, 2026—a co-founder of one of the most powerful companies on the planet sitting next to the Bishop of Rome on the anniversary of the most important social encyclical in Catholic history—is a scene of the redefinition of the West’s cultural power. Not only because the Vatican has chosen to side with the Silicon Valley faction that the Trump White House has excluded from its military contracts. But because, in doing so, it has agreed to perform a legitimizing function that every system of power requires and seeks: the function of translating economic and technical dominance into recognized moral authority.
The question that remains open—and which the history of the coming decades will have to answer—is whether the Vatican is truly attempting to govern the artificial intelligence revolution through the autonomous force of its own moral tradition, or whether it is becoming an integral part of it: not the judge of the system, but its priest. Not the prophet who speaks to power, but the master of ceremonies who consecrates it.
In 1891, Leo XIII had paid the price for his autonomy: Rerum Novarum had displeased Catholic capitalists as much as atheist socialists, and neither camp had embraced it with enthusiasm. It remained an uncomfortable document, capable of disturbing all the comfortable certainties of its time. Will Magnifica Humanitas be capable of the same discomfort? Will it be able to question the system instead of legitimizing its more moderate protagonists? Will it know how to ask the question that none of the parties involved wants to hear: to whom does the future that artificial intelligence is building belong, and under what conditions will its wealth be distributed?
These are questions that the May 25 ceremony raised without answering. And perhaps it is in this silence, rather than in the official words, that the true meaning of the encounter between the Vatican and the algorithm lies. The issue is not merely about technology. It concerns who controls its symbolic and moral meaning. And whoever controls the meaning, ultimately, controls the future.