Every technological revolution arrives with its own apologetics. It promises efficiency, abundance, and relief from drudgery, and often delivers a measure of all three. What it does not deliver is the moral vocabulary required to govern it. Work changes faster than judgment. Power concentrates faster than law can restrain it. Households become vulnerable before anyone has counted the cost.
This state of affairs describes the world into which Pope Leo XIII sent Rerum Novarum — “Of New Things” — on May 15, 1891. The encyclical described “the spirit of revolutionary change” that had passed from politics into economics. The “new things” were the industrial factory, the wage-labor contract, and a concentration of wealth that had disfigured the social fabric within a generation.
In our own day, Rerum Novarum continues to provide a model for engaging the AI revolution, which risks outrunning the moral vocabulary needed to govern it.
Given that Pope Leo XIV sees his ministry as a continuation of the tradition Leo XIII inaugurated 135 years ago this week, the anniversary of this foundational text of modern Catholic social teaching is an urgent invitation to reflect on AI in explicitly Leonine terms.
Man Before the Machine
Most readers likely remember Rerum Novarum as the encyclical that endorsed workers’ associations and rejected socialism. This is true. But a deeper, anthropological argument runs underneath those conclusions.
Leo XIII’s claim was that the human person, endowed with the capacity for reason, does not live by appetite alone. We plan, we provide, we assume responsibility across time. That is why private property matters: It is the stable material condition of foresight, household stewardship, and intergenerational provision. Work should give the worker access to that stability, not sever him from it.
Property, on this account, is the durable form that labor’s fruits take for the sake of the household. Family is where those fruits pass, in love, to those who did not earn them — and from the household’s surplus, charity reaches the wider community. Strip these capacities away, or reduce them to transactional form, and you have not merely an economic problem but a wounded person.
The 1891 crisis, as Leo diagnosed it, was that industrial modernity had replaced these capacities rather than amplified them. The worker had become, in his words, a “mere instrument for money-making,” valued solely for his physical powers. The factory did not abolish the human; it displaced the human from the acts that make one human.
Amplification or Replacement?
Leo XIII was emphatically not a Luddite. He did not propose to dismantle industry. He simply insisted that any new technology be measured against the person it was meant to serve.
Applied to AI, the question becomes: Does the tool amplify or replace human capacities? Does this new arrangement extend the worker’s personal capacities, or absorb them? Does the tool advance her skill, or make it irrelevant? The distinction is mine, but it presses a recognizably Leonine concern.
I have argued in these pages that this is precisely the question we must ask of AI in the classroom: Does the tool amplify the student’s capacities — judgment, discernment, responsibility for truth — or replace them? An AI that drafts a student’s essay while he remains incapable of defending his thesis has not educated him; it has produced on his behalf. The goal of education is not the production of essays but the formation of a person with a lively, responsible mind.
The same test now applies beyond factories and schools: to offices, hospitals, courtrooms, firms, and increasingly the home itself. The technology is morally complex; it can amplify, and it can replace. The question is whether we will have the clarity and the courage to tell the difference.
Application to AI
Consider four areas where the test clarifies the stakes, with the family as the foundation beneath them all.
First, judgment. Industrial machinery displaced the body’s labor, but the worker still owned the judgment that ordered his work. AI threatens something the steam engine did not: displacement not only from an act, but from judgment about the act. Judgment is not merely the verdict that concludes thought. It is the operation by which a person frames a question, imagines alternatives, weighs what is at stake, and stands behind the answer. A workplace in which judgment is outsourced to a model leaves the professional, tradesman or technician with tasks to perform but not acts to own. Does the tool amplify the worker’s skill or reduce her to just one cog in a mostly impersonal system?
Second, the fruits of labor. Leo XIII argued that labor establishes a just claim to what it produces. Pius XI later clarified in Quadragesimo Anno that production involves the combined efforts of labor and capital, so the fruits cannot be assigned to labor alone. But that still leaves a serious question when models are trained on the aggregated work of writers, coders, illustrators and analysts whose contributions may be unattributed and uncompensated. The magisterium has not given a final judgment; the tradition nonetheless tells us what must be weighed: Does the new tool amplify or reduce consent, attribution, compensation and the common good?
Third, concentration. Leo XIII named, without apology, “a small number of very rich men” who had laid a yoke on the laboring masses. Our analogue is not hard to see: a handful of firms control the foundation models, computing power, and data on which contemporary AI depends. Markets and ownership are legitimate, but they must serve the common good, including the spiritual good of persons, not merely the material good. Does the tool amplify or reduce a just distribution of capital and power?
Fourth, association. Leo XIII did not imagine the worker as a solitary contractor standing naked before power. He defended “workingmen’s associations”: communities of formation, mutual aid, and shared responsibility, not merely instruments of bargaining.
That principle transfers directly to the AI economy. When hiring, evaluation, promotion and dismissal are mediated by opaque systems, workers need institutions that preserve contestability and human review: professional societies, faculty governance bodies, medical boards, and other mediating institutions through which judgment is formed and held accountable. Does the tool amplify or replace the particular human good of association?
Association, for Leo XIII, is a natural right — and its purpose is not the pursuit of advantage against the employer, but the protection of the person against impersonal power of any kind.
Preeminence of the Family
Underlying the workplace is the family, which the Church consistently teaches is prior in nature and rights to the state; Leo XIII treated it as the deepest beneficiary of a just economy. A father worked so his children would not suffer want; a mother’s work anchored the household; inheritance carried labor’s fruits into the next generation. This was not sentimentality but social architecture — the foundation on which everything else rested.
AI’s economic effects are now reaching that foundational level. The disruption is not confined within professions; the ability of people to make a livelihood and support their family, now and in future generations, is insecure.
Leo XIII’s question about the just wage — whether work returns to the worker enough to sustain a household — is exactly the question an AI-era economy will have to answer. When stable careers fragment, when middle-class knowledge work becomes precarious, when young adults cannot project a future stable enough to marry and have children, the displacement is not merely individual. It is generational, and our evaluation of AI’s effects must take it into account.
Continuing the Leonine Legacy
The Catholic social tradition did not end with Rerum Novarum. St. John Paul II’s Laborem Exercens distinguished between the objective dimension of work — what is produced — and its subjective dimension: the person who works and is formed through working. The amplification-or-replacement question pays attention to this subjective meaning of work. Benedict XVI’s Caritas in Veritate likewise reminded us that technology is never merely technology: Every tool carries within it a conception of the human person and must be placed within a moral horizon, rather than permitted to become the horizon.
Pope Leo XIV has already made the continuity explicit. In his May 10, 2025, address to the College of Cardinals, he explained that he chose the name Leo because the Church now faces another social and economic revolution, which poses “new challenges for the defense of human dignity, justice and labor.” In his June 2025 message on AI ethics, drawing on Antiqua et Nova, he insisted that access to data must not be confused with intelligence and that AI must be judged by integral human development. In a December 2025 address to labor consultants, he put the point in the simplest Leonine terms: Work must not revolve around capital, market laws or profit, but “the person, the family, and their well-being.”
Leo XIII insisted that no purely procedural solution can reach the social question; only religion and morality can do that. The AI question cannot be addressed apart from religion and morality, either. Procedure alone, however efficient or well-informed, cannot produce what persons and institutions must become.
The “new things” of any age must not determine the human ones. The person cannot be algorithmized away any more than she could be industrialized away. The Church’s call for discernment is constant, regardless of new technologies, because the person it defends does not change.
What Leo XIV is saying fits within a framework that is 135 years old this week. Rerum Novarum remains the document most worth rereading as Leo XIV continues to develop the Church’s reflection on artificial intelligence.
