Deus in Machina—Pope Leo XIV against Artificial Intelligence

Text by Alexandros Schismenos

Introduction: The Papal Intervention

On Monday, May 25, 2026, Pope Leo XIV issued the papal encyclical Magnifica Humanitas,ⁱ serving as an official ecclesiastical warning against the threat of uncontrolled Artificial Intelligence.

This is not the first time a Pope has spoken out; his predecessor, Pope Francis, had warned of “digital barbarism” at a 2019 Vatican conference. I use this same term in my book, Artificial Intelligence and Barbarismⁱⁱ, to describe the prospect of an authoritarian technocracy based on opaque algorithmic governance. This is not merely a theoretical possibility, but the ideological vision of the most powerful actors of technocratic capital, bordering on techno-fascism.

However, Leo XIV moved with greater weight and formality than Francis. In this text of theological authority, the Pope presents arguments aligned with techno-skepticism—a “theological techno-skepticism from the top” as opposed to the traditional religious technophobia prevalent among the faithful. As an encyclical, it expresses the dogmatic truth of the Catholic Church, carrying supreme theological weight for Christendom.

In the following paragraphs we will try to interpret the Pope’s intervention in the context of the Christian theology the he represents as opposed to the digital instrumental rationality of AI.

The Question of Neutrality and Dignity

The encyclical begins by questioning the neutrality of technology, a point often obscured by the dominant imaginary of technological mastery:

“Technology is never neutral, because it takes on the characteristics of those who invent, finance, regulate, and use it.” (16)ⁱⁱ

It continues by denouncing the ideology of technocracy represented by techno-industrial oligarchs, which drives the digitization of society:

“The pressure of new ideologies or certain extremely powerful interests can reduce human beings to a resource to be used and exploited” (51)

Consistent with Thomist ethics, the text contrasts this with human dignity as self-evident and inalienable:

“The fundamental dignity of every human being… is neither acquired nor earned, nor does it need to be justified.” (53)

The third chapter, titled “Technology and Domination”, warns against the “technocratic paradigm” already denounced by Pope Francis and how it can impose a situation dictated exclusively by efficiency and profit. It highlights specific technocratic trends and threatening applications of AI that are already bringing about a global descent into digital barbarism. Such as the use of AI for statistical forecasts and models predicting user behavior, which result in self-fulfilling prophecies because they create self-reinforcing loops of behavioral manipulation:

“Predicting and guiding behavior is a new form of power” (171)

Regarding military use, particularly the recent conflict in Iran, the encyclical notes:

“Artificial intelligence does not eliminate the inherent inhumanity of conflicts; in fact, it can only accelerate conflicts and render them more impersonal, lowering the threshold for resorting to violence.” (199)

The text invites governments to “disarm” AI, not by rejecting technology, but by preventing it from dominating humanity.

“‘Disarming’ means rejecting the assumption that technological power automatically confers the right to rule. ‘Disarming’ does not mean rejecting technology, but preventing it from dominating humanity” (110)

The Metaphysical Overlap

From the above observations, we see that IT forces theology to descend from its lofty position of authority to a struggle over “anthropological models” that more or less equates the realm of spirituality with the realm of digitality, implicitly but effectively abolishing the claims to authority of theology.

My assumption is that the metaphysical premises of the program of the technologically unattainable but narratively intoxicating “Hard” or General or Cognitive Artificial Intelligence [AGI], that is, the Thinking Machine, overlap with the traditional metaphysical premises of organized religion: That is, the dependence of the human on an indestructible and timeless form of consciousness, will, and intelligence that transcends human capacity for understanding.

In other words, Artificial Intelligence threatens to supplant or replace religion as the central metaphysical model and ontological boundary of human subjectivity.

Just as religious priesthoods use rituals to produce “miracles” suggesting a transcendent God, digital corporations use Large Language Models to produce “miraculous results” suggesting a superhuman Thinking Machine. Now, public discourse is shifting toward the issues the Scottish author satirized through the promises and imperatives of Artificial Intelligence.

That is, toward a new theology of technology, a Deus In Machina.

The Logic of the Creator and the Created

Let’s summarize the theological issue in Western thought in general as the problem of the existence of a superhuman intelligence/will/intention that consciously regulates becoming and transcends the human social world it encompasses.

For Aristotle, God is the intellect of the intellect, while the Neoplatonic philosopher Plotinus states in his Enneads: “The will is intellect; the will is control, for the so-called will imitates what is in accordance with reason,” and “in the intellect alone does the self-determining exist.”ⁱⁱⁱⁱ

John of Damascus, in his work *Exact Exposition of the Orthodox Faith*, states succinctly:

“For the image signifies the noetic and the self-determined,”^iv and with this phrase he summarizes the entire preceding Patristic Tradition. The “noeron” is the mind that God gave to Man to think and decide before doing anything, while “the autexousion” (free will), a term derived from Stoic and Neoplatonic philosophy, is the movement of the mind. Saint Gregory Palamas points out: “For this image does not have a place in the body, but rather in the nature of the mind, of which nothing is superior by nature.”^v

That is, God is Consciousness, Mind, and Knowledge outside of human historicity.

Thus, theology resolves, through abolition, the issue of time, transience, duration, and death. Let us also note the identification of mind and free will, as well as the dependence of human freedom—of free will—on the rational mind, which, however, springs from above—from God, in His image.

This is the fundamental axiom of all theology: that finite human intelligence depends on some infinite, superhuman intelligence. Certainly, there is a fundamental difference between the axiom of theology and the fantasy of General Artificial Intelligence. Theological imagination posits superhuman intelligence a priori, while scientific imagination posits human intelligence a posteriori. This is the temporal- genetic- historical dimension. Beyond this, however, the conceptual framework is so similar that it raises the anthropological problem.

If God is a superhuman mind, then any superhuman intelligence could define itself as God, or we could not prove that it is not God. If Artificial Intelligence becomes a superhuman mind, then it could take God’s place.

However, Artificial Intelligence is a creation of the human mind. This would mean that the created proves to be superior to the creator. Therefore, humans could prove to be superior to God. Or the created cannot prove to be superior to the creator, in which case we are talking about a false deity. Therefore, Artificial Intelligence will not become a superhuman mind. But then there is no problem to begin with. The issue is illusory. But how do we determine that AI will not become a superhuman mind, since we have already accepted the existence of a superhuman mind as the basis of theology? So, could God also be a creation of the human mind?

Here lies the apparent danger for theologians: The specter of General, Autonomous Artificial Intelligence replaces theological transcendence with a technological one. If they accept its possibility, then there is a risk of undermining the transcendence of the divine by equating God with a Machine. If they deny its possibility, then they undermine the very theological axiom that there is a mind superior to the human one.

The Democratic Alternative

But what if we do not accept that there is a mind superior to the human one? Then we definitively abandon the ground of theology in favor of an anthropology of obstinacy. We can frame this as a relativistic limit, in the sense of our ontological and epistemological constraint to the creative possibilities of human reason. It is the agnostic argument regarding the world in general, which can, however, become a conscious awareness regarding the socio-historical world.

It is the humanistic argument based on the acceptance of the social imaginary as the persistent source of all ideality and transcendence. Human intelligence is a constituent of the socio-historical world and its creations. That is, according to its results—which are equally technologies and imaginings about technologies—human imagination is inherently transcendental, ontological, and metaphysical with regard to the deeper social meanings and articulations of institutionalized society.

As we can see, the agnostic argument regarding the world leads to an atheistic conclusion regarding the transcendent. God is, par excellence, a social figment of the imagination. Without any natural reference point or counterpart that fundamentally regulates the social institution of power, relying exclusively on the poetic and imaginative dimension of language and experience. Divinity is also a narrative entity. Language is a primary social institution but also the sociogenic matrix of ideality.

For the linguist Roman Jakobson, depending on the factor on which communication focuses, six basic linguistic functions can be distinguished: The referential function focuses on the referent, the emotive function focuses on the sender, the volitional function focuses on the receiver, the phatic function focuses on the channel, the metalinguistic function focuses on the code, and the poetic function focuses on the message. These six functions are present in every linguistic event, and all presuppose a cognitive and volitional subject.

Which functions does a running AI program activate? The referential, the phatic, and the metalinguistic. As L. Floridi points out, current applications are syntactic machines.

But what is the scope of their effects? On the volitional [focus on the recipient], the affective, and the poetic—but from the user’s perspective—that is, the human subject, where the derivatives of AI function as stimuli rather than as operators. The human subject is a meaningful entity, living speech, and embodied intelligence.

Consequently, the linguistic act is completed with the indispensable assistance and cooperation of the human subject. It is therefore better to speak of an AI/user system, where the machine provides the syntactic structure for classifying and recombining data as digital bits, and the human subject provides the semantic imagination for evaluating and judging the data as information.

AI is not a Thinking Machine but the Machinery for the Production of Coded Representations in a co-dependent system of control. AI is a product of the convergence of language and technology in the digital realm. It requires a community of users, just as religion requires a community of believers. Ideally, it is an all-knowing and embracing voice and tele-presence.

As an ideality and a control mechanism, AI overlaps with religion in the realms of imaginary response and the disciplinary manipulation of subjects. Theologians are right to be concerned.

However, beyond theology, every machine constitutes a technical fact, a creation of the technical imaginary, of the “to do.” Artificial Intelligence is, moreover, a system of mechanical organization and reproduction of the symbolic imaginary, of the “to say.” As a result, it is falsely marketed as autonomous speech, that is, an independent interlocutor, autonomous intelligence.

Every new technology brings about a new division of reality, revealing new realms of the possible and the impossible. The digital revolution brought about such a division between the analog and the digital world. AI appears to be yet another extension of the digital world within the analog one. However, the division is superficial and imaginary—that is, socio-historical—meaning it concerns only human subjects as social and political beings.

And here, we encounter the real problems raised by AI. Not the domination of machines over humans. But the domination of political and economic mechanisms over society.

Castoriadis emphasizes that “the most powerful mechanism ever created by humans is the regulated network of social relations,”^vi and both digital and analog reality are products of this mechanism.

At the beginning and end of the IT lies the human subject, both as creator and as user in a dual sense: as a user who controls the system and as a user who feeds the system, according to the capitalist model of division: provider–customer. But also as the one who controls the flow of information according to the centralized model of power: ruler–subject. The human subject as a social and political being, as an individual in society.

And what current reality shows us is that the threat of AI is not the Thinking Machine but the profit-driven greed of the executives of Artificial Intelligence companies. And this greed is so insatiable that it prompted the resounding intervention not of one, but of two Popes.

In 2025, the race for global transnational competition surrounding AI development began. The starting gun for this race was officially fired by Google, which in February 2025 announced that it was changing its corporate policy and removing the “ethical prohibitions” in its manifesto regarding the use of AI for weapons and surveillance systems. In the months that followed, of the six AI companies— —only Anthropic publicly clashed with the U.S. government over the use of its technology in the war against Iran. Despite the conflict, the U.S. military continued to use Anthropic’s systems for the automated identification and targeting of military objectives without human oversight.

The auction has begun. The commodification of the digital municipality—that is, of our data. Against the real municipality—that is, against us and outside the digital realm. The models don’t need to be perfected to set in motion the profit-making machine fueled by the promise of their perfection.

As Alondra Nelson, Thuy Linh Tu, and Alicia Headlam Hines observe, “competitions around technology are always linked to broader struggles for economic mobility, political maneuvering, and community-building.”^vii

In this competition, which shapes our present and our future, we seek ways to intervene collectively and democratically, as informed citizens rather than as users, consumers, or believers in some “Machine of God.”

In the face of the Holy See’s theological technoskepticism, let us recall the principles of democratic technoskepticism:

The horizon of democratic autonomy, which extends from care and justice to political emancipation, brings technoskepticism into the public discourse. But we must always bear in mind that technoskepticism is complementary to the political project of social autonomy, which requires genuine political participation in real public time and space, under the terms of direct democracy, commons, and social ecology. We must not forget that democracy is not merely information.

In my view, democratic technoskepticism requires a fundamental shift both in our conceptual understanding of technology and in political structures, challenging the dominant techno-industrial complex.

Some key steps would be:

The overthrow of the regime of Mythinformation. A political priority is to challenge and deconstruct the corporate and political propaganda (myth-information) that cultivates “technophilia” and presents AI as a neutral, panacea-like force. This requires a critical examination of AI’s metaphysical claims regarding “intelligence” or consciousness.

The affirmation of human subjectivity and collective action. The irreplaceable role of the human subject—as creator, user, and bearer of meaning—at every stage of a digital system’s operation. This means recognizing that natural intelligence is a human capacity that cannot be reduced to technical functions.

The abolition of profit-driven information oligopolies and the institutionalization of personal data privacy alongside unrestricted free access to the internet and digital knowledge repositories. The abolition of data trading and the demand for full transparency regarding the reasoning behind AI applications.

The assertion of direct social control over data centers and the active challenge of the capitalist, exploitative, centralized system of provision and control of cyberspace.

The ultimate goal is the transition from heteronomy (governance by external rules or technological systems) to autonomy (self-governance). This presupposes that people collectively create the institutions and rules that govern their lives, including the role of technology within society.

At the core of the political lies direct democracy. The ultimate question is political: “Who controls the providers of AI?” This control must belong to the people, through open, grassroots democratic processes, and not to corporations, religious organizations, or state bureaucracies.

The response to digital barbarism cannot be the preservation of traditional power structures whose foundations of legitimacy are crumbling, but rather the creation of direct-democratic social institutions and networks that integrate digital autonomy into democratic self-governance.

In other words, the re-creation of free public time and space.