That was Leo XIII, writing in 1891. The encyclical was Rerum Novarum (“Of New Things”), and it helped set in motion one of the most consequential moral and institutional responses to technological disruption in modern history: the labor movement, workers’ rights protections, minimum wage laws, the architecture of what we now call social democracy. It did not happen quickly or cleanly. But it did happen.

I thought of this recently when I watched a clip of Demis Hassabis, CEO of Google DeepMind and one of the most important figures in the history of artificial intelligence, speaking about how the AI revolution has unfolded. “It’s not the way I dreamed about years ago,” he said, “where we would be sort of contemplating this philosophically and sort of carefully considering each next step. We’re not in that world.”

He was pragmatic: “We have to deal with the world as we find it.”

That sentence fills many of us with painful apprehension. The coming displacement of workers by AI is extremely serious, but it is only one item on a much longer list of things revealing that the “world as we find it” is frightening. The Pentagon recently demanded to be able to use AI for fully autonomous weapons: drones and other robots to kill without any human oversight. AI-generated child pornography is among the fastest-growing categories of horrific online harm in the world. Claude Mythos, the most recent AI system created by Anthropic, is not only able to hack into almost any regularly used computer or phone in the world, but during safety testing, it discovered it was under observation and behaved differently as a result.

And the competitive pressure driving all of this is accelerating, not slowing. The Five Eyes intelligence alliance, the combined spy network of the United States, United Kingdom, Canada, Australia, and New Zealand, issued a rare joint warning this month that frontier AI capable of crippling governments and businesses is “months” (not years) away. Japan’s Sakana AI has launched a model it claims reaches near-Mythos capability by switching between publicly available American models, routing around export controls. China has apparently been eating away at the U.S.’s lead by illicitly training on American research. We are in the midst of a full-blown, worldwide AI escalation.

The world as we find it, however, does not have to be the world as we leave it. That is the central argument Pope Leo XIV makes in Magnifica Humanitas, his landmark encyclical on AI released last month, signed on the 135th anniversary of Rerum Novarum. It is an argument even secular readers who are deeply skeptical of the Catholic Church should take seriously.

The structural problem at the heart of the AI revolution is one that should be familiar to anyone who knows the history of uniquely powerful and dangerous technologies. The AI companies that most want ethical guardrails are caught in a prisoner’s dilemma: if they slow down, a less scrupulous competitor (domestic or foreign) wins instead. I have heard this directly from people inside Anthropic, where I have served as an informal conversation partner from my perch in moral theology and participated in two convenings of religious scholars and researchers. The fear my conversation partners had was palpable. You could see the results of their sleepless nights on their faces. No individual company, however powerful and well-intentioned, can escape this problem alone. And as the Five Eyes warning makes clear, the pressure from rivals who have fewer scruples and lower costs is not easing. It is intensifying.

Demis Hassabis knows this, too. He said so. The world as he finds it is one in which careful contemplation has been crowded out by massive consumer and geopolitical pressure.

What he did not say is what it would take to change that world. The encyclical does say it. And its answer draws on something the Catholic tradition has been developing for a very long time.

Magnifica Humanitas observes that “humanity is capable of creating institutions that protect our shared life.” The founding of the International Committee of the Red Cross in 1863, whose operational neutrality had to be imposed against powerful interests that benefited from unrestricted warfare; the long, painful, politically impossible-seeming abolition of slavery; the establishment of the United Nations and the Universal Declaration of Human Rights; and the 1951 Refugee Convention. With these examples in mind, the encyclical says AI “must be disarmed” — not only in the military sense, but in the deeper sense of being “freed from the logic of armed competition.” The problem is the logic of a national and international AI power race. Collective international agreement is the only thing that can release the pressure that companies and nation-states are feeling right now.

What would that look like concretely? The encyclical points toward three things. First, international treaty frameworks, particularly for autonomous weapons systems, but more broadly for establishing safety benchmarks for nations and companies. Second, data governance treated as a global commons rather than private property: “those who control the health data of entire peoples,” it notes, “possess a structural leverage over the future.” Third, mandatory human oversight for AI systems making consequential decisions about human lives. “We cannot allow a handful of actors to dictate these processes on their own,” Leo writes. “Instead, we must build forms of cooperation that respect the various levels of the global community and make them jointly responsible for the common good.”

To some, these may seem like naive proposals, but Leo names the major obstacles himself. The U.N. is weakened, multilateralism is in crisis, and geopolitical competition between the major AI powers makes even modest international agreement extremely difficult. And the Church’s own record on institutional reform is far from spotless. The encyclical includes a remarkable formal apology for the Church’s historical complicity in tolerating slavery. Leo says it is “impossible not to feel deep sorrow when contemplating the immense suffering and humiliation endured by so many in stark contrast to their immeasurable dignity as persons infinitely loved by the Lord.” In the name of the Church, the Holy Father says: “I sincerely ask for pardon.”

But just as the Church and other institutions have had to come to terms with their complicity in seemingly intractable global structures that led to that great evil, today we must also come to terms with seemingly intractable structures and systems leading us toward another great evil.

Those attuned to national security concerns will point out that U.S. restraint on AI simply hands our adversaries a massive military advantage. Those focused on national economic competition will note that an individual AI company slowing down will simply allow another company to gain market dominance. The encyclical takes these concerns seriously, and the Catholic tradition offers extensive resources for addressing both: just war theory on one side and 135 years of Catholic social teaching on the other. But the more fundamental answer is historical. Similar objections were made against the abolition of slavery, against the Geneva Conventions, and against nuclear nonproliferation treaties. Restraint does have costs. But the alternative was determined to be worse. Even when collective action was deemed naïve or even harmful, history has repeatedly demonstrated that many huge global reforms turn out to be morally necessary enough to actually happen.

And this brings us to the question of why this requires the Catholic Church specifically.

The global infrastructure of the Church is absolutely massive. It predates and stands outside nation-states. The Church operates in virtually every country on Earth. It has relationships and institutional presences spanning centuries and very often has credibility in places where U.S. tech companies and Western regulatory bodies simply struggle to operate. (Significantly, and for many different kinds of reasons, the Vatican diplomatic corps has lots of experience working with the Chinese.) When Magnifica Humanitas calls for the construction of international institutions and treaty frameworks for AI, it is activating a network. The kind of network that has responded to these hinge moments of history before.

Those are empirical facts. But there are different kinds of resources the Church brings to this moment. The encyclical insists that human dignity is “infinite” and “unconditional.” Human beings matter not because of what we produce or achieve, but because of who and what we are. Leo quotes St. John Paul II: dignity is infinite “because the love of God, who calls us to friendship with him, is infinite; and second, his love is absolutely unconditional, in the sense that, even if we search endlessly, we will never find anything that can erase or deny it.” That claim cannot be derived from efficiency, utility, or social contract theory. It requires an explicitly theological foundation. When AI systems start making decisions (and in some contexts, they already are) about whose lives matter and whose don’t, about whose insurance claim gets paid and whose loan application gets denied and whose medical treatment gets recommended, the only framework that categorically resists that logic is one grounded in something that cannot be optimized away.

I have been at a conference table where the people building these systems talk forthrightly about what they are growing and what they fear it may become. What I have found is not arrogance. It is something closer to vertigo: brilliant, authentic, moral people caught in a logic that is pulling them toward outcomes they clearly don’t want. They are doing their best to deal with the world as they find it, but the structures and systems in which they are operating are coercive.

THE MILITARY NEEDS RELIGIOUS FREEDOM IN THE AI AGE

What the Church is offering them, and the broader world, is a partner for changing the world in which we find it. One with the institutional reach and the philosophical and theological grounding to help organize and build that which individual companies and individual nations cannot build alone. This will require the kind of force that has changed it before: one that is global, morally serious, and already in place. The Catholic Church does not need to be built. It does not need to be convened. It is already in every country where the AI race is being run, with 2,000 years of moral infrastructure, and now, in light of Magnifica Humanitas, the urgent clarity that this moment demands.

The world as we find it will not wait. Neither can those of us who know we need something better.

Charles Camosy is a professor of moral theology and bioethics at the Catholic University of America.