Magnifica Humanitas

In the [YOU] on AI Field Guide

The cycle's foundational metaphor—AI as an amplifier that carries whatever signal you feed it, at civilisational scale—finds its institutional counterpart in the encyclical's single most-quoted line: technology takes on the character of those who build and use it. The amplifier does not evaluate the signal; the moral weight lies entirely with the person feeding it. The Pope and Edo Segal are saying the same thing in two dialects. Where [YOU] on AI asks 'Are you worth amplifying?', the encyclical asks 'Which city are you building?'. Both questions point to the same excavation: the quality of the human being at the centre of the machine.

The encyclical's treatment of concentrated private power adds rigour to one of the cycle's sharpest warnings. A handful of firms control training data, model weights, and platform distribution for billions of people. The document names this with unusual bluntness: private technological actors now hold resources and capacities that 'surpass those of many Governments.' The cycle traces how the river of intelligence risks enclosure by the few who can afford the dam; the encyclical calls the same concentration a structural violation of the universal destination of goods—a principle the Church had previously applied only to land, water, and physical resources.

The document's meditation on work gives the cycle's account of professional identity disruption its deepest moral register. When the Church insists that the subjective dimension of work—who the worker becomes through the labour—is primary over the objective dimension (what is produced), it supplies the frame within which the automation of cognitive work is not merely an economic event but a violation of the conditions under which a person finds meaning. This is Viktor Frankl's argument made institutional.

Origin

The encyclical was released on 25 May 2026, ten days after signing, set deliberately against the 135th anniversary of Rerum Novarum. Leo XIV chose the name Leo precisely to signal this lineage. Its immediate precursor, the shorter note Antiqua et Nova (2025), had already established the foundational distinction the encyclical would develop: machine intelligence is powerful information processing; human intelligence is something wider, bound up with embodiment, conscience, and the search for meaning. Magnifica Humanitas is the full-dress elaboration of that distinction across an introduction, five chapters, and a conclusion, citing more than two hundred sources.

The launch was itself an argument about who should be in the room. Alongside Cardinals Fernández and Czerny and the theologians Anna Rowlands and Léocadie Lushombo sat Anthropic co-founder Chris Olah, the first non-cleric to speak at such a presentation. The Holy See simultaneously established an interdicasterial commission on AI drawing together seven departments, signalling a sustained institutional commitment rather than a one-time statement. The document's release in eight modern languages simultaneously, with no Latin version, was another deliberate departure: the Church was speaking in the present tense, to the present world.

One of the document's most disarming moves is a formal apology for the Church's historical complicity in slavery, citing Leo XIII's 1888 letter In Plurimis and acknowledging that 'it took eighteen centuries' to recognise the full incompatibility of slavery with the Gospel. This confession is not a detour; it is the institution warning the builders of AI that the same mechanism—the steady pressure of political and economic convenience overriding moral clarity—operates in every age, and that no civilisation should be confident it is not currently blind to an injustice it will later have to repent.

Key Ideas

Babel and Jerusalem. The encyclical's central image is not a technical metaphor but a civic one. Babel is built by a crowd seeking self-aggrandisement, erasing difference in the name of unity, ending in the collapse of shared speech. Jerusalem under Nehemiah is rebuilt by families each tending the stretch of wall nearest their own home, coordinated by shared purpose rather than by a master plan. The question the encyclical poses on its first page is which of these two cities the AI industry is currently building. The Babel syndrome, as the document calls it, names the specific pathology: 'the idolatry of profit that sacrifices the weak, uniformity that neutralizes differences, and the illusion that a single language can translate the mystery of the person into data.'

Infinite dignity. The load-bearing claim is that every human person possesses infinite dignity, inalienably grounded in existence itself, that 'is neither acquired nor earned, nor does it need to be justified' and that holds 'in and beyond every circumstance.' Against the machine's structural tendency to represent persons as rankable vectors, the encyclical plants a counter-claim: infinities do not sit on leaderboards. A civilisation that grades human worth by usefulness has wagered everything on a race it is now losing to its own tools. The antidote is not a better algorithm but a prior conviction that resides in being rather than output.

Universal destination of goods, extended. Catholic social teaching has long held that the earth's resources were given for the whole human family. The encyclical's decisive extension is to digital goods: 'Among the goods that are universally intended for everyone, we must also include new forms of property, such as patents, algorithms, digital platforms, technological infrastructure and data.' When these concentrate without adequate sharing, 'a new imbalance is created.' The document is explicit: 'Ownership of data cannot be left solely in private hands but must be appropriately regulated.' This is the same logic that once governed common land, now applied to the invisible commons of the digital age.

Subsidiarity as AI governance. The principle that decisions should be made at the lowest effective level, and that higher authorities support rather than swallow lower ones, becomes in the encyclical the foundational criterion for evaluating AI deployment. Communities 'must not be reduced to passive recipients of decisions made elsewhere.' This requires transparency in algorithms, equitable access to data, avenues for recourse, and structures of genuine participation—not the democracy of users who can choose between platforms but the Jerusalem model of people who actually help build the wall.