The Pope's AI Encyclical Isn't Really About AI

In Brief

Pope Leo XIV's first encyclical, Magnifica Humanitas, released May 25, 2026, frames artificial intelligence as an anthropological crisis rather than a technical one. The 82-page document calls for the disarming of AI and warns the technology could make civilization less human.

Anthropic co-founder Christopher Olah appeared alongside the Pope at the Vatican presentation, marking unprecedented industry-Church dialogue. The encyclical rejects just war theory as outdated and includes a historic Vatican apology for the transatlantic slave trade.

Cardinal Michael Czerny explicitly stated: The encyclical is not about AI. It's about the human person.

The question of what it means to remain human in an age of intelligent machines is precisely what Human x AI Europe will examine on May 19 in Vienna, where policymakers, technologists, and ethicists are gathering to work through these tensions together.

When the Vatican announced that an AI company co-founder would stand beside the Pope to present a major Church teaching, the pairing seemed incongruous. The Catholic Church warning about technology while sharing a stage with someone building it? The apparent contradiction dissolves once the actual argument comes into focus.

Cardinal Michael Czerny, presenting the document, offered the clarifying frame: Please note that the encyclical is not about AI. It's about the human person.

This distinction matters for anyone trying to understand what the Vatican is actually saying, and what it isn't.

The Disagreement Beneath the Document

Magnifica Humanitas arrives at a moment when AI debates tend to split along predictable lines. On one side: those who see the technology as an unprecedented tool for human flourishing. On the other: those who see it as an existential threat requiring immediate constraint. The encyclical refuses both framings.

Pope Leo XIV, the first American-born pope, deliberately chose his papal name to invoke Leo XIII, whose 1891 encyclical Rerum Novarum guided Catholic social teaching through the Industrial Revolution. The parallel is intentional: just as that earlier document addressed not machinery itself but the human conditions machinery created, this one addresses not algorithms but what algorithms reveal about human choices.

The document identifies what it calls an anthropological crisis. This is a specific claim worth unpacking. An anthropological crisis is not a technical problem requiring technical solutions. It is a crisis about what kind of beings humans understand themselves to be, and what kind of society they are building.

When Leo writes that AI risks making civilization less human, he is not making a prediction about job displacement or algorithmic bias, though the document addresses both. He is making a claim about meaning: that systems optimized for efficiency and data can hollow out the practices through which humans recognize their own dignity.

Where the Positions Actually Diverge

The encyclical's call to disarm AI has generated headlines, but the term requires disaggregation. Does disarming mean:

• Prohibiting autonomous weapons systems?
• Constraining AI development for geopolitical dominance?
• Limiting commercial applications that reduce human agency?
• Rejecting the framing of AI development as an arms race?

The document appears to mean all four, but these are distinct policy positions with different implications. Someone might agree with the first while finding the fourth naive. Someone might embrace the third while considering the second strategically unworkable.

The encyclical explicitly warns against a race for ever more powerful algorithms and larger datasets, driven by the desire to secure geopolitical or commercial dominance. This is a values claim, not a technical assessment. It asserts that the competitive dynamics driving AI development are themselves morally problematic, regardless of what the technology produces.

This is where the document's argument becomes genuinely challenging for policymakers. It is not simply calling for better guardrails on existing development trajectories. It is questioning whether those trajectories are compatible with human flourishing.

The Anthropic Presence

Christopher Olah's appearance at the Vatican presentation deserves attention beyond its symbolic novelty. Olah leads Anthropic's interpretability team, the group working to understand what happens inside AI systems when they process information. This is not a public relations role. It is a technical research position focused on one of the field's hardest problems: making AI systems legible to human understanding.

The choice of Olah specifically, rather than a CEO or policy executive, signals something about the nature of the dialogue the Vatican is seeking. Interpretability research asks: can humans understand what these systems are doing? This question has obvious resonance with a document concerned about human agency and dignity.

Anthropic has organized multiple meetings with religious leaders, including a March 2026 gathering at their San Francisco headquarters where Christian scholars received briefings on how large language models are built and trained. Meghan Sullivan, a philosopher at Notre Dame who attended, described the company as really invested in these dialogues.

Brian Patrick Green, director of technology ethics at Santa Clara University, offered an assessment: Of all the people in the world who work on AI that they could have chosen from industry, he's probably the right person.

This framing suggests the Vatican is not simply seeking industry validation. It is seeking interlocutors who take the interpretability problem seriously, who acknowledge that the systems being built are not fully understood by their creators.

The Just War Rejection

The encyclical's rejection of just war theory as outdated has received less attention than its AI passages, but the two arguments are connected. Pope Leo has repeatedly clashed with the Trump administration over the Iran conflict and religious justifications offered for it.

The document states: Today, more than ever, without prejudice to the right to self-defense in the strictest sense, it is important to reaffirm that the 'just war' theory, which has all too often been used to justify any kind of war, is now outdated.

This is not a tangent from the AI discussion. Autonomous weapons, AI-enabled targeting systems, and algorithmic warfare are precisely the contexts where just war reasoning meets technological capability. The encyclical's rejection of the framework applies directly to debates about human oversight in military AI applications.

The document notes that the U.S. Department of Defense has pursued legal action to ensure it can develop autonomous weapons which kill without any human oversight. The encyclical treats this not as a regulatory question but as a moral one about what humans are willing to delegate to machines.

What the Document Asks

Near its conclusion, Magnifica Humanitas anticipates reader overwhelm. Pope Leo quotes Gandalf from The Lord of the Rings: It is not our part to master all the tides of the world, but to do what is in us for the succor of those years wherein we are set, uprooting the evil in the fields that we know, so that those who live after may have clean earth to till.

The literary reference is not decorative. It addresses a specific objection: that individual choices cannot matter against systemic technological forces. The encyclical's response is that this sense of powerlessness is itself a subtle temptation to be resisted.

For policymakers and technologists, the document poses a question that cannot be answered with technical specifications or regulatory frameworks alone: What does it mean to build systems that serve human dignity rather than merely human efficiency?

The encyclical does not provide a policy roadmap. It provides a frame for evaluating whether the roadmaps being proposed are asking the right questions.

Whether one accepts the Vatican's authority on these matters is beside the point. The argument stands or falls on its own terms: that the AI debate, properly understood, is not about AI at all. It is about what humans are choosing to become.