A note on method before beginning. The encyclical runs to two hundred and forty-five paragraphs, and the word “dignity” recurs in it one hundred and one times. This is not a stylistic tic: it is the keystone of the entire edifice. And keystones, in a piece of reasoning, are exactly the points that must be tested first – because if that one gives way, all the rest gives way too.

First crack

The whole document rests on two contrasting biblical images: the tower of Babel and the rebuilding of Jerusalem under Nehemiah. On one side the proud construction, the uniformity that flattens, the dominion that dehumanises; on the other the shared labour, each with their own stretch of wall, the communion. Leo XIV says it explicitly in paragraph 9: “the first choice is not between a 'yes' or a 'no' to technology, but between building Babel or rebuilding Jerusalem.”

It is a powerful image, and it works beautifully as rhetoric. As logic, it is a false dilemma – the most classic of fallacies: presenting two options as exhaustive when others exist. But the real trap is not in the bifurcation itself. It is in how Babel is defined. In paragraph 7 we read that the tower is “a work conceived without reference to God.” Keep that sentence in mind, because it does all the dirty work. If Babel is by definition the project built without God, then any secular collective enterprise – any immanent, horizontal, atheist attempt to build a more just world – falls into Babel not for its outcomes, but for its premise. The die is loaded before the throw. We are not choosing between dominion and fraternity: we are choosing between “with God” and “without God,” dressed up as an ethical choice.

The problem is that Nehemiah's Jerusalem refutes the encyclical itself. Reread paragraph 8: the city is reborn “through the shared responsibility of the whole people: priests, artisans, heads of household, women and the young,” each with their own piece of wall, listening to fears, coordinating efforts. Strip away the theological frame and it is the exact description of mutualism. It is bottom-up self-organisation, mutual aid, the federated cooperation that anyone who has spent time with libertarian thought recognises at first glance. The “way of Nehemiah” does not need the Lord at the centre to function: it needs people who trust each other enough to entrust each other a stretch of wall apiece. Which is exactly what the free software communities, the decentralised networks, the commons projects do – without praying to anyone. The encyclical describes the model, admires how it works, and then insists that without God that model would be Babel. It does not prove it. It postulates it.

There is a third city, the one the text erases by definition instead of by argument: the city built together, from below, with no heaven to reach and no one to ask for permission. It happens to be the city some of us have been trying to build for a while now.

Second crack

Let us come to the keystone, the one with the hundred and one occurrences. The anthropological pivot of the encyclical is the concept of “ontological dignity,” set out in paragraphs 52 and 53. It is the dignity that belongs to every human being “simply by the fact of existing, of having been willed, created and loved by God.” A dignity called “infinite” – taking up the 2024 declaration Dignitas infinita – because “infinite is the love of God that calls him to friendship with Him.”

Let us stop on the structure of the argument, because it is a perfect circle. Human beings have infinite dignity; dignity is infinite because God loves them with infinite love; and we know it holds for everyone because everyone is created by God. And there it is: the conclusion is inside the premise. It is a textbook begging of the question: to accept the foundation you must already have accepted exactly what the foundation is supposed to demonstrate, namely the existence and the love of that God. For a believer it is coherent, and that is fine. But the document addresses itself – paragraph 16 – “to all men and women of good will,” claiming a universal validity that its own foundation denies it.

Here the encyclical makes a clever move, and it must be acknowledged. In paragraph 56 it reverses the charge: without a solid metaphysical foundation, it says, human rights become negotiable, and “rights held today to be untouchable may, in the future, end up being called into question or denied by those who hold power.” It is a consequentialist argument – if there is no God, rights collapse – used to prop up a thesis meant instead to be deontological. But it is doubly fragile.

First: it is itself an argument about consequences, not about foundation; it is telling me it is in my interest to believe, not that it is true. Second, and more important: it is simply false that without metaphysics rights are left without ground beneath their feet. Contractualism, the ethics of reciprocity, moral consideration grounded in the capacity to feel and to suffer – all found dignity on intersubjective and verifiable bases, with no need to posit a creator. One can debate which is best. But they exist, they work, and they hold.

The encyclical, however, has already thought about how to neutralise them. In paragraph 133 whoever grounds their values on human reason alone is described as “modern man wrongly convinced of being the sole author of himself,” a victim of “a presumption, consequent upon a selfish closing-in on oneself.” Translated: the secular refutation is not refuted, it is diagnosed as the sin of pride. It is an elegant way of not having to answer.

One could object that the encyclical, elsewhere, grants reason the ability to get there on its own: in paragraph 56 it admits that reason, questioning itself on human nature, “is able to discover values that hold for all.” But it is a poisoned concession. That “human nature” with its objective values already inscribed within is not a neutral secular datum: it is itself a metaphysical construction, natural law presenting itself as self-evidence. And there is worse, because reason is admitted to discover those values only if it arrives at the right conclusion. When it does not get there – when it grounds dignity on bases of its own, without a creator – paragraph 133 kicks in and that same reason becomes “presumption.” You have permission to reason, provided you reason as they do. It is not a shared foundation: it is a confessional foundation with a service door that closes the moment you try to leave it by another way.

Third crack

The technical description in paragraphs 98 and 99 is surprisingly accurate. The idea that modern models are “more 'cultivated' than 'built,'” that developers “create an architecture on which the AI grows,” and that “fundamental scientific aspects – such as the internal representations and computational processes of these systems – remain at present unknown” is simply true, and it is the language of mechanistic interpretability, not of theology. On this, no objection: it is the most honest thing in the document.

The problem comes immediately after, when the correct description is used for an incorrect move. In paragraph 99 it is established that one must “avoid the misunderstanding of equating this 'intelligence' with the human one,” because the systems “do not live an experience, do not possess a body, do not mature in relationship,” and above all “do not understand what they produce.” So far it is a defensible definition. But it is deployed to express a sealed paradigm: whatever a machine does, however sophisticated, “will never be true intelligence” because it lacks the “affective, relational and spiritual” horizon. Every counterexample is excluded by redefining the term in a way that makes it inaccessible by construction. The capacity to compute is there, the sophistication is there, the utility is there – but the soul is not, and the soul is precisely what it had been decided from the start the machine could not have. The conclusion was already in the definitions. Note that I am not claiming the models are conscious: I am saying that an argument that makes its own thesis unfalsifiable is not an argument, it is a definition in disguise.

There is then a cost this move makes the text pay, and it is the most serious. Those verbs – to understand, to know, to create – the encyclical uses as if their meaning were fixed and settled, on one side the machine that does not deserve them, on the other the human who possesses them by right. But that is exactly what today is no longer settled. These systems are putting under pressure the paradigms with which we define knowledge, creativity and relationship, and the scientists who study them know it perfectly well. Whether “to understand” means anything for a machine is an open question: some argue that to predict the next word accurately a form of understanding has to be built, and there is the Othello-GPT experiment – a model trained only on game transcripts, never on the rules, that internally developed spontaneously a representation of the board.

On creativity the confusion is almost comic: one psychometric test places the models in the top one per cent for originality, another finds them lacking precisely in originality – a sign that we do not even know how to measure the boundary we claim to draw. And that the machines are changing the way we think is said by the research on cognitive offloading: those who delegate more to AI show lower critical thinking capacities. On this the encyclical is right, in paragraph 100 it says it almost in the same words. But it is precisely here that the text bites its own tail: a stochastic parrot does not raise the question of what it means to understand – these systems do. It is a phenomenon that forces us to rethink what knowledge and creativity are, and a question of that kind is not dispatched with a definition taken for granted.

Fourth crack

The most refined fallacy of the whole text sits in the paragraphs running from 126 to 128. The encyclical confronts transhumanism – the promise of a technical overcoming of human limits – and opposes to it its own “more than human”: grace, the elevation worked by God in Christ, the “transcending of oneself” that “surpasses the capacity of nature.” I quote 128: “we come to be fully human when we are more than human, when we allow God to lead us beyond ourselves.”

It is an equivocation, in the technical sense: the same label – “more than human” – is applied to two radically incommensurable things, and the two things are then presented as if they competed in the same market. On one side technological enhancement, measurable, in principle verifiable. On the other the transformation by grace, which is an act of faith unverifiable by definition. The encyclical disqualifies the first as Promethean illusion – the self-sufficiency that mimics salvation – and proposes the second as the “authentic” transcendence. But for someone standing outside the enclosure of faith, they are two claims of the same kind: two promises of overcoming, neither of the two demonstrated.

And it is not a forcing of mine to set the two things in competition: it is the text that does it. It is the encyclical that reuses the same formula – “more than human” – for grace; it is the encyclical that presents the desire for overcoming intercepted by transhumanism as an authentic thirst to which only divine transcendence would give the “true” answer. Once you have established that the need is the same and that only one of the two offers is legitimate, the competition is one you have declared yourself.

What remains for me is only to note that the two offers are not of the same order: one promises something measurable, the other something that can only be believed. To treat one as fantasy and the other as reality is not the result of an argument. It is the presupposition from which the argument starts: I convince you that the opponent's pseudo-transcendence is empty, and meanwhile I sell you mine as full.

Fifth crack

Paragraphs 118 to 120 contain the most poetic passage of the encyclical, and it is precisely for this the most insidious. The thesis is that “the human does not flourish in spite of the limit, but often through the limit”: illness, old age, vulnerability, suffering are not defects to be corrected but places in which the human matures.

There is an undeniable psychological truth here: sometimes from suffering wisdom is born, from failure a growth. But observe the slippage. We start from a descriptive claim – from suffering value sometimes derives – and we land on a normative one: therefore to reduce the limit technically is hubris, it is the “purely technical salvation” to be rejected. It is the naturalistic fallacy, but inverted: from the fact that finitude can generate good, it is deduced that intervening to attenuate it is morally suspect. But “suffering sometimes teaches” in no way implies “therefore we must not fight it.” From the observation of a fact no duty follows.

And in paragraph 120 there is the gem that says it all: “to suppress pain entirely one would have, at bottom, to switch off love and desire too.” To the credit of the text, it must be said that the encyclical does not at all deny the duty to heal: the same paragraph 118 acknowledges that “it is a duty to seek to eliminate suffering.” The point then is not whether to intervene, but where the boundary lies between legitimate intervention and the “overreach” to be condemned. And that boundary the encyclical draws without giving us a criterion: aspirin yes, hybridisation no, but in between? Who decides when healing becomes desecrating? Between aspirin and the uploading of the mind into a cloud there is a continuum as long as a life – anaesthesia, vaccines, prostheses, antidepressants, glasses – and without a non-arbitrary criterion that boundary remains a decision, not a deduction. It is worth recalling, in passing, that it is exactly the logic of the “natural limit not to be desecrated” that has historically been used against obstetric anaesthesia, contraception, assisted reproduction. Every time, today's limit was sacred until someone crossed it; and the day after no one dreamed of putting it back into question.

Sixth crack

This is less a single fallacy than the load-bearing structure of two entire chapters. The first and second chapters – from paragraph 28 to 89 – are a long chain of references: Leo XIII, Pius XI, Pius XII, John XXIII, Paul VI, John Paul II, Benedict XVI, Francis. Every thesis is anchored to a predecessor, in a line that feeds itself.

It must be said with precision, because here it is easy to miss the mark: inside the Catholic system, the continuity of the Magisterium is not a fallacy, it is the criterion. For a believer, the fact that a doctrine has been coherently held by eight pontiffs across some hundred and fifty years is a legitimate argument, because the authority of that tradition is an accepted premise. The problem arises at the exact moment the encyclical leaves the enclosure and addresses everyone – believers and not – claiming universal validity. There the argumentum ad verecundiam becomes visible: the chain of citations counts as proof only for those who recognise the authority of the chain. For everyone else it is a circle closing on itself, imposing as you like, but self-referential. Eight popes agreeing with one another do not constitute a proof for anyone who recognises in none of the eight the right to pronounce.

Water to one's own mill

There remains the question I was posing at the start: is it simply “bringing water to the Church's mill,” or is there something more?

It is water to the mill in the structural sense of the term, and you can see it with the naked eye once you have isolated six (there are probably more) cracks. The scheme is always the same, repeated chapter after chapter: a shareable secular diagnosis – technocratic power, exploitation, algorithmic dehumanisation – channelled towards a non-negotiable confessional therapy. A diagnosis I would sign, but a solution I cannot accept without first accepting the theological premise. The fifth chapter, on war, repeats verbatim the binary structure of the introduction: “culture of power” against “civilisation of love,” Babel against Jerusalem under other names. The rhetorical machine is the same, oiled and tireless.

But it would be short-sighted to stop here, and I refuse to do so for three reasons.

The first is that the diagnosis is solid, and as such it makes the encyclical a tactical ally. When Leo XIV writes, in paragraph 108, that “small, very influential groups can orient information and consumption, condition democratic processes and bear on economic dynamics to their own advantage,” he is saying something true and saying it from the most listened-to pulpit on the planet. When, in 109, he recognises “the invisible, often exploited, labour that feeds the algorithmic models” – the data labellers of the Global South paid a pittance to train our chatbots – he is doing materialist critique, not catechism. On this terrain we are on the same side of the barricade.

The second is that there is at least one point where the encyclical leaves its own mill and applies its principles to itself. In paragraph 89 there is talk of “listening to the victims of spiritual, economic, institutional, sexual abuse, of power, of conscience” within the Church, with “the recognition of the harm, the just reparation and the prevention.” It is little, it is late, and it is written in the velvet language of the Curia. But it is dialectically honest: it is not water to the mill to admit one's own structures of sin, and it is right to give credit for it – while knowing that one entry in a list is not yet a reckoning, and I will return to this at the end.

The third is that behind the document there is a real operation of power, and not at all naive. In a global regulatory vacuum on AI – where states limp along and civil society struggles to find a voice – the Church puts itself forward to fill the space as a planetary moral authority. It does so, intelligently, by lining up against the private power of Big Tech, which makes it attractive to anyone who criticises that power. It is, in its way, a textbook move: occupy a terrain that others have left undefended. That it is then a terrain on which we too would like to build something – a collective governance of technology, data as commons, slowing down where everything accelerates – is precisely what makes the encyclical so slippery. It agrees with you on the destination for five-sixths of the journey, and then at the last fork it turns one way only.

Conclusions

Every fallacy I have listed, taken on its own, could be a stumble. Taken together, they design a method. They are not the errors of sloppy reasoning: they are the devices of an extremely careful reasoning, and each one performs the same precise function. They serve to make a diagnosis the secular reader shares converge towards a conclusion that, without the premise about God, they would have no obligation to accept. The false dilemma closes off the alternatives from the start; the begging of the question on dignity makes faith the only admissible foundation; the equivocation on “more than human” disqualifies every competing transcendence; the naturalistic fallacy turns the limit into a duty; the appeal to authority closes the circle. Take away God, and the argument does not hold – and it is built on purpose so that you, to make it hold, must put God back in.

This does not make it a bad document. It makes it a partisan document pretending not to be one, which is a different thing. Magnifica Humanitas is excellent sociology, magnificent rhetoric, and logic that limps exactly – and only – at the points where the supernatural has to be let in. The technical part on AI, the one written with the contribution of those who actually study the models, is the most solid and the least ecclesial. The anthropological part, the one on which all the rest rests, is the most fragile. It is no accident: it is where the text has to do the work it cares about most.

As an atheist who shares half the premises and none of the conclusions, the same question remains that I ask myself every time someone describes to me a just city and then explains that I cannot build it without their god. Quoting Eric Raymond, I have seen the Bazaar – and not the Cathedral – work. I have seen it work in free software, in the networks that have no master, in the communities that hold themselves together through mutual aid. I have seen it work even without a god. The question I leave open, then, is simple: if we raise the wall just the same, each with our own piece, listening to one another and trusting one another – who said there must necessarily be someone up there at the centre? And what if we noticed it held up perfectly well without?

A small postscript: someone will object that the encyclical does have courage – it asks forgiveness for the delay with which the Church condemned slavery. True. But it is the most comfortable forgiveness there is – for a fault of eighteen centuries ago, which touches no living bishop. On the abuses of today there remains one line, one entry in a list of six in paragraph 89. This is why I insist: lining up against techno-capitalism, artificial intelligence and transhumanism is today a very intelligent social and political positioning to take, but really a very uncourageous one. If I may advise Leo XIV something truly courageous, let him try to write an entire encyclical against paedophilia in the Catholic Church. I will gladly offer myself as first reader.

Sources and further reading

The document – Leo XIV (2026). Encyclical Letter Magnifica Humanitas on the guardianship of the human person in the time of artificial intelligence. Full text. – Dicastery for the Doctrine of the Faith (2024). Declaration Dignitas infinita on human dignity.

The tradition invoked (to find your bearings in the magisterial chain) – Leo XIII (1891). Rerum novarum. – Second Vatican Council (1965). Pastoral Constitution Gaudium et spes. – Francis (2015). Laudato si'. – Francis (2020). Fratelli tutti.

On logical fallacies (general references) – Walton, D. (2008). Informal Logic: A Pragmatic Approach. Cambridge University Press. On the false dilemma, the argumentum ad verecundiam and the slippery slope as argumentation schemes and on their legitimate and illegitimate uses. – Hume, D. (1739). A Treatise of Human Nature, book III. The classical formulation of the distinction between is and ought, at the root of the so-called naturalistic fallacy.

Secular foundations of dignity and rights (the alternatives the encyclical pre-disqualifies) – Rawls, J. (1971). A Theory of Justice. Harvard University Press. The contractualist foundation of justice without metaphysical presuppositions. – Singer, P. (1979). Practical Ethics. Cambridge University Press. Moral consideration grounded in the capacity to feel and to suffer. – Nussbaum, M. (2006). Frontiers of Justice. Harvard University Press. The capabilities approach as a basis for human dignity.

On AI interpretability (the most solid technical core of the document) – Olah, C. et al. (2020). “Zoom In: An Introduction to Circuits”. Distill. On why the internal representations of models remain largely unknown even to those who build them.

On the knowledge / understanding debate in LLMs (the paradigm revision) – Bender, E. M., Gebru, T. et al. (2021). “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?“. FAccT '21. The text that coins the “stochastic parrot” metaphor: a system trained on form alone cannot access meaning. – Li, K. et al. (2023). “Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Task”. ICLR. The Othello-GPT experiment: a model that internally builds a representation of the board without ever having seen its rules. – Tayyar Madabushi, H., Torgbi, M., Bonial, C. (2025). “Neither Stochastic Parroting nor AGI: LLMs Solve Tasks through Context-Directed Extrapolation”. The middle position: capacities that go beyond the parrot but remain predictable and not assimilable to human cognition.

On computational creativity (and on how uncertain the boundary is) – Boden, M. (2004). The Creative Mind: Myths and Mechanisms, 2nd ed. Routledge. The founding distinction between combinatorial, exploratory and transformational creativity. – Guzik, E. et al. (2023). “The originality of machines: AI takes the Torrance Test”. Journal of Creativity. GPT-4 in the top 1% for originality and fluency. – Lu, Y. et al. (2025). “Assessing and Understanding Creativity in Large Language Models”. Machine Intelligence Research. The opposite result: LLMs excel in elaboration but are lacking precisely in originality.

On AI's impact on human cognition – Gerlich, M. (2025). “AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking”. Societies, 15(1): 6. Negative correlation between frequent AI use and critical thinking, mediated by cognitive offloading. (The author warns: correlation, not causation.)

On power embedded in technological choices – Winner, L. (1980). “Do Artifacts Have Politics?”. Daedalus, 109(1): 121-136.

#MagnificaHumanitas #LeoXIV #Encyclical #AI #Atheism #LogicalFallacies #FreeSoftware #Commons #Mutualism #Philosophy #Writing