AI Gets Religion
Early reflections on the Vatican's new AI encyclical: Magnifica Humanitas
What does a 2000-year-old world religion have to say about powerful AI? It turns out, a lot – over 42,000 words give or take, depending on the language in which you are reading the new encyclical. Most of the news in the coming days will focus on the optics of the rollout – Anthropic’s leadership being on the dais, there will inevitably be speculation about tensions with the Trump administration. But the real star this week should be the text itself, which is one of the clearest presentations of Catholic social teaching in recent memory and which inaugurates Pope Leo as a powerful and compelling voice in AI ethics.
Magnifica Humanitas opens by explaining how the Church continually develops its social teaching to meet new issues; Pope Leo offers case studies illustrating the ways in which the Church aims to be a place where each generation comes for ethical discernment. This isn’t bluster. Last fall, I attended a closed-door meeting on AI ethics at the Vatican with a group of scholars, theologians, and prominent figures in the American technology industry. The meeting began with an optional early morning Mass at an ancient church in Rome. Seated a few pews away from me at the service was one of the tech leaders, the kind of guy you typically see in a black t-shirt and chinos. That morning he was dressed in a brown suit and tie, quietly taking in the sanctuary as the first rays of morning light filled the room.
After the service, I pulled him aside. I hadn’t known he was Catholic.
“I’m not,” he said.
“Then why are you here?”
He paused. “We’re building something that is going to change life as we know it. I want to make sure I keep in touch with what humans have always cared about. This is a place that takes care of those values.”
While leaders will still make the pilgrimage to Rome for moral guidance, we now have an authoritative document to guide our day-to-day thinking about AI. We will need the weeks ahead to pore over Magnifica Humanitas. I come at it first as a philosopher who also leads a $50.8m initiative to provide practical resources for putting Christian teachings about human dignity and AI into action. My first read of this beautiful document yielded three “can’t miss” themes.
First, the deepest challenge AI poses is anthropological. It forces us to ask: What, exactly, makes human beings special? For most of modern history, the implicit answer has been our minds. In particular, our capacity for abstract reasoning, creative expression, and decision making. But now, commercial software on our phones can write poetry, pass bar exams, diagnose diseases, and generate artistic images. The implicit answer turns out to be false. This realisation has the ability to undermine all of our ethical thinking, which has long rested on the foundation that human life has profound value and should be upheld.
Pope Leo uses the encyclical to forcefully argue that the Church has a better theory of where our dignity lies. Christian tradition has never grounded human dignity in cognitive performance or economic productivity. It has never said: you matter because of what you can do. Rather, the faith says that you matter because of who you are. Someone with a body, mind, and soul. Someone built for love. Someone with an intellect oriented toward truth and will that can be held accountable. We are vulnerable in a way that AI models are not. And Pope Leo argues that this special belovedness and vulnerability makes us magnificent. He writes: “Finitude, when truly accepted, does not diminish us but opens us to recognizing the face of God and others. Indeed, precisely because we experience limits – vulnerability, suffering and failure – we can recognise the inviolable dignity of every person, both our own and that of others” (122).
This leads to the second major theme I picked up in the encyclical. Pope Leo offers a blistering critique of the technocratic paradigm that goes unquestioned in much of Western culture. In this, he continues a project that was central to Pope Francis’s ethical teaching in Laudato si’. Technocratic ideology views every decision as an occasion to realise efficiency, gain further control, and extract a profit.This is often the logic of the market. Pope Leo points out how worried we should be that a small number of private companies rather than democratic governments are now making the most weighty decisions about which AIs to develop and release. Reflect on just how much of your own mindshare is now dominated by news about these companies and their products. Pope Leo writes: “When [technology] becomes the standard by which everything is judged, it begins to dictate what matters and what can be discarded, reducing creation to an object of exploitation and human beings to mere cogs in a system driven toward ever greater efficiency” (92).
Perhaps the most insidious forms of technocracy arise when we think this is the way we must think about ethics. There are major schools of thought in technology companies, at universities, and in government who believe our most profound ethical obligation is to work to optimise the human condition using technology. They believe we have an obligation to saturate our institutions with AI – that we have an ethical obligation to enhance human intelligence using AI. They assert that we have an obligation to minimise disabilities and to plan ways to enhance the abilities of distant generations. On this approach, all ethical reasoning boils down to a cost-benefit calculation, and a small number of engineers and economists control everyone else.
The alternative to the technocratic ethic is an ethic of love. Pope Leo makes extensive use of an idea coined first by St. Augustine and later made into a Church teaching by Saint Paul IV – the ethical ideal for humanity is a civilisation of love, “a social order in which justice and charity are intertwined and love becomes the guiding principle of economic, political and cultural life” (186). For Pope Leo, any progress we make with AI needs to be measured against this standard: Is this helping the human family grow in love and solidarity with one another? The first twenty years or so of the digital revolution have largely failed this test. Social media has fed toxic political hatred. Small numbers of people have become very rich while new underclasses have been created. And while we now have unprecedented access to information, we resolutely choose what is convenient over what is true. Ethical progress with AI will only occur when these trends are arrested and reversed.
This leads to the final, “can’t miss” element of this encyclical: the practical guidance Pope Leo offers to policymakers and everyday folks on how to think about AI in light of this ethical vision. We need to take a global perspective on this technology and we have to resist the shallow realpolitik which has taken hold among world leaders. While global leaders like Donald Trump and Xi Jinping frame these debates about who will “win”, the US or China, there are 6.5 billion other people on planet earth whose lives are being upended by this technology. Pope Leo argues that the Church must be a voice for them, especially the poor, who are increasingly excluded from the benefits of this technology and forced into desperate situations, to mine the rare earths we need for computer chips, to devote their days to manually filtering the (often horrifying) visual data used to train machines, and to allow large, extractive data centres to be built near their homes.
Pope Leo also offers a clear message to corporate leaders on the rollout of AI across the workforce. Unemployment is a grave evil, especially when caused by profit-driven automation. Work is how we contribute to the common good, and it is a key element in human flourishing. He argues that we need social criteria for judging AI innovation, we need continuous support for re-skilling and professional transitions for workers, and we need to hold corporations accountable for their commitment to the human dignity of their employees (156).
And Pope Leo exhorts schools and universities, who have the unenviable task of preparing a generation of human beings to be natives with this powerful technology. Educational institutions must remain the primary place where human knowledge is united and authoritatively advanced. And more than ever they must be places of human formation: “Schools are not called to follow the pace of the digital world, but to offer that which the digital sphere by itself cannot provide, namely a shared time for learning and developing trustworthy relationships” (147).
We have our work cut out for us supporting advances in how we teach human dignity that can keep pace with the relentless onslaught of educational software. At Notre Dame, with the DELTA Network, we are taking on this challenge by building a large network of K-12 educators committed to formation-first approaches to AI in education. And in the coming weeks we will announce a major grant for over fifty college professors committed to building and launching courses on AI ethics that are rooted in human dignity.
A teaching like this is obviously a cause for celebration among Catholics. But people of all faiths will hear their deepest concerns about AI represented by Pope Leo XIV. More importantly, they will see a roadmap for how we should now envision the world we should build with this technology in it. Magnificent indeed.
