Whose Dominion? - American Reformer

Pope Leo, the Journal, and the Question Neither Will Ask

When Leo XIV released Magnifica Humanitas, his first encyclical and a 43,000-word meditation on artificial intelligence, the reaction divided along the lines you would expect. The Wall Street Journal’s editorial board welcomed his defense of human dignity in a May 28 editorial, but rejected his “faith in the state,” warning that his call for regulation echoed Bernie Sanders and Alexandria Ocasio-Cortez. The progressive left, predictably, welcomed exactly what the Journal feared.

Both sides are having the wrong argument.

The Journal and the Pope are quarreling over the administrator — should AI be governed by the state, the market, or the international community? — while sharing, without examination, the same underlying assumption: that artificial intelligence exists to serve a single, universal humanity. The Pope calls it the “common good” realized in the “community of love.” The Journal calls it the free market, operated responsibly. Both want a different steward for the same humanist project.

Scripture frames the matter differently, and the difference is the whole point. This essay is a Protestant reading of Magnifica Humanitas — an attempt to show that the question AI forces upon us is not more state or freer market, but “Whose dominion?”, and in whose hands a tool of such power should rest.

The Humanism Beneath the Social Doctrine

For all the hype, the encyclical is only nominally about AI. The bulk of it is a restatement of Roman Catholic Social Doctrine — common good, subsidiarity, solidarity, social justice, integral human development — and the anthropology underneath it. Leo takes a long time to get to the point, and when he does, the point is made in humanistic language, not covenantal language.

Consider subsidiarity – the principle that decisions should be made at the level closest to the persons involved. It is an inherent contradiction within the Roman system. You cannot claim that authority should rest at the lowest possible level while retaining a man in Rome who claims plenary, universal, immediate jurisdiction over every baptized soul on the planet. The contradiction is not incidental; it is baked in, and Rome cannot resolve it from within.

Joshua Mitchell, in these pages (“Whither the Reformation in America?”), has named the deeper reason. Rome’s subsidiarity descends from Aristotle by way of Thomas — a telos-bound metaphysics in which grace is superadded to nature. The Reformation made a different wager: not Athens but Jerusalem, not nature but covenant. Subsidiarity and federalism look alike at the surface — both devolve authority — but they do entirely different metaphysical work underneath. Federalism rests on covenant. Subsidiarity rests on a doctrine of man that does not survive Genesis 3. Grace cannot build successfully on a deformed nature.

This is why Leo’s solidarity reads as pseudo-covenantal. “No one is saved alone,” he writes — borrowing the cadence of federal headship while quietly removing the covenant that gives it meaning. It sounds biblical, but it is parasitic. We are not, in fact, all brothers and sisters. There is no one big human family. There are God’s Elect and the reprobate, and any social doctrine that begins by erasing that distinction ends somewhere other than the Gospel.

AI as the Test Case

When Leo turns to AI itself, his observations are often sound. He sees that AI is not morally neutral. This is correct, and ironically makes AI more human, not less, since every tool inherits the moral fingerprints of its maker. Less concretely, he also sees that power is concentrated in the hands of a few and worries about the handover of responsibility to systems no one can answer for.

But Leo never examines the implications of that handover. I argued in “Crowns in the Gutter” that the real danger is not “the AI” but the owners of the AI — or, more precisely, whoever writes the prompts the AI executes. This is Leo’s worry spelled out, but the Pontiff never spots the opportunity. When Leo writes that injustice arises “almost automatically” from “structures, mechanisms and economic and cultural systems,” he acknowledges what critics long suspected: that critical theories and their applications, like DEI, now govern Catholic Social Doctrine.

If injustice is structural, and AI is trained on those structures, then AI inherits a kind of original sin from its human progenitors. Therefore, the structures must be dismantled. Decolonized. Leo writes, “Even today, colonialism assumes new forms. It no longer dominates only bodies, but appropriates data, transforming personal lives into exploitable information.”

The decisive moment comes when Leo demands that AI be subjected to “shared standards of social justice.” Leo now places all his cards on the table. This is the criterion for earning Leo’s imprimatur: adherence to “shared standards” that are neither shared nor just. It is a power grab: the use of the Pope’s office to remove a tool of dominion from the very men who need it most. Town councils defending home rule against “affordable housing” mandates. Embattled conservative pastors and congregations in historic denominations captured by progressives. AI promises to level the playing field for the little guy. Leo wants to make sure the little guy remains a subsidiary.

The Civic Vacuum

Leo sees the handover of power happening globally, but the parish council, the town committee, the school board, and the zoning commission are the real sites where the question “who writes the prompts?” will be answered. Operational authority in American civic life is about to migrate, not to AI itself, but to whoever directs it at every level of government.

This migration could, ironically, realize the very ends Catholic Social Doctrine has chased after for a century — but realized in federalism, not subsidiarity. The people in the room, at the meeting, write the prompts. Not the unelected city manager. And certainly not a bishop in Rome. The answer to Leo’s anxiety about neutrality is not to engineer a neutrality that cannot exist, but to place the tool in accountable hands. Covenant men. This is how AI becomes a tool of dominion rather than domination — a tool of governance under God.

And this is where the Journal’s answer fails just as surely as the Pope’s. The invisible hand cannot say to whom it is accountable. It relocates power without naming a master. The market is not neutral either; it simply hands the prompt-writing to whoever has the most capital. Neither a beneficent state nor an unfettered market can supply what a covenant can: men answerable to God for what they build.

Inheritance, Not “Universal Destination”

The theological core of Leo’s poorly framed encyclical surfaces in his favorite phrase, which appears in the encyclical more than a dozen times: the “universal destination of goods.” The teaching has patristic roots — Chrysostom wrote that to withhold wealth from the poor is to rob them — and it is consistent with Rome’s brand of universalism, urbi et orbi, the city and the world. Rome is the imperial center; her clients, subsidiaries. “The poor you always have with you,” Jesus said in John 12:8.

The Bible does not speak of the universal destination of goods. The Bible speaks of an inheritance. Christians stand to inherit the earth with Christ. If goods are destined for anyone, they are destined for Christ and for Christians. Covenant-keeping men and women can expect a portion of that inheritance even now, in this life: “There is no one who has left house or brothers or sisters or mother or father or children or lands, for my sake and for the gospel, who will not receive a hundredfold now in this time” (Mark 10:30).

Artificial intelligence has the potential to realize precisely this kind of multiplication. Where Leo sees a good to be redistributed by subsidiaries directed by a universal administrator, the covenant sees an inheritance to be stewarded by a covenant people. Leo’s humanism is squeamish when it comes to accepting the terms of covenant nationalism.

The Rabble Are Promised Nothing

Leo ends his encyclical where he began it: with two biblical cities. Babel, the tower built to make a name for man apart from God, and Jerusalem, rebuilt by Nehemiah through the shared labor of a covenant people. Leo casts his lot, rightly, with Jerusalem — “men and women prepared to enter the construction sites of history… to rebuild what has collapsed and protect what is threatened.”

It is the right image. But Jerusalem was not rebuilt by humanity in the abstract. It was rebuilt by a covenant people, called out and set apart, each assigned his portion of the wall, who, after they had finished the wall, publicly confessed God and “separated themselves from all foreigners” (Nehemiah 9:2).

Those are not words that are safe for Leo to say. Behind them, Leo fears, lurk the strong gods AI might summon. His fears are not unfounded, but his faith in the New Jerusalem — the elect lady — ought to sustain him more than it does. He frets over the rabble, the children of Hagar, born of the flesh. The promise is not to them.

Neither Will Say His Name

Man, the Psalmist says, is made “a little less than God” and crowned “with glory and honor.” Is it possible that AI, made by man to be a little less than man, might be a jewel in that crown — not a rival to be feared, nor an idol to be worshipped, but a tool fit for the dominion God gave Adam and restored in Christ?

That is a question neither the Pope nor the Journal can answer, because both have already decided that the steward of AI must be a universal one — the Curia, the United Nations, or an autonomous one — the market. The Protestant answer is older and more unfashionable: not a universal, undifferentiated humanity enthralled by the imperial center, but a covenant people, answerable to God, building under His name in the place He has put them.

The Pope and the Journal both want a safe steward for artificial intelligence. Neither will say His name.