Pope Leo's AI Warning Is Not About AI

Pope Leo XIV chose a particular moment to speak about artificial intelligence. The Vatican released his first encyclical on 25 May 2026 — the same week that Deutsche Welle announced it was closing its accounts on X, platforms were moderating content ahead of yet another political season, and the European Union's AI Act moved into its enforcement phase. The timing was not accidental. The encyclical is framed as a technological warning. It is, at closer reading, an institutional one.

The document — titled, according to early excerpts, with a direct call to "disarm" the technology — urged governments to slow deployment and increase regulatory oversight. It used language about misinformation, democratic erosion, and the concentration of decision-making power in the hands of a narrow technical and financial elite. None of these concerns are novel. What distinguishes the encyclical is the vector through which they are sequenced: the Pope is arguing that AI is not primarily a technical problem awaiting a technical fix. It is a governance problem for which technology has been volunteered as a proxy.

This framing puts the document at odds with the dominant conversation in Washington and Brussels, where AI regulation is largely treated as an exercise in calibrating the pace of deployment against the risk of catastrophic output. Tech companies prefer the narrow framing because it positions them as the only parties with the expertise to manage the variables—effectively making them both the subject of regulation and its most credible interpreters. Governments have often accepted this premise, sometimes enthusiastically. The encyclical rejects it.

A Traditional Diagnosis in Contemporary Dress

Those familiar with the Vatican's longer political tradition will find the encyclical's structure familiar: enumerate the harms, locate their source in a structural concentration of power, recommend that power be checked by institutional rather than market mechanisms. That sequence has defined much of the Church's economic and political teaching since Rerum Novarum. What is new is the technology in the middle of it.

Applied to AI, the same logic produces an argument that sounds, to many Western observers, like Luddism. It is not. The encyclical does not reject the technology. It rejects the governance gap around it—the absence of meaningfulchecks on who builds these systems, on what data they are trained, and on whose interests they are designed to serve. The critique, in plain terms, is not that AI is too powerful but that too much of it is controlled by too few actors who face too little meaningful accountability.

TechCrunch's read of the document was more precise than most: "Pope Leo XIV's first encyclical uses AI as a lens to diagnose older problems: concentrated power, eroding democracy, and a tech elite that shapes the world to its own advantage." That is the short version. The encyclical is conservative in the structural sense—its argument is that institutions designed to constrain power have not kept pace with the systems that concentrate it, and that technology, left unmanaged, will not correct for that.

The Power the Document Actually Describes

The specific actors the encyclical targets are not hard to identify. A small number of firms control the训练的算力和模型基础设施 that everything else depends on. They shape regulatory conversations through lobbying and through what their spokespeople choose to call acceptable risk. They have, in several documented cases, moved faster than governments understood what was being deployed. The encyclical treats this not as a failure of individual corporate judgment but as a structural feature of a system in which governments have ceded territory they have not recovered.

What is interesting about the document is that it does not offer a solution that belongs to any established political faction. It is too skeptical of both market self-correction and state centralization to sit comfortably inside either progressive or conservative frameworks. It calls for what amounts to a pluralization of AI governance—multiple institutions with real oversight authority, including bodies whose mandate is broader than economic performance. That is a harder ask than most regulatory proposals because it requires not just new rules but new institutional architecture.

The Polymarket framing — "Pope Leo XIV calls to 'disarm' AI" — is technically accurate but flattening. Disarm suggests a weapon. The encyclical is more careful than that. It describes a set of relationships—between capital and infrastructure, between technical elites and democratic accountability, between what technology makes possible and what governance permits—and argues that those relationships need to change. The weapon metaphor flatters the technology by suggesting it is the active agent. The document's logic is that the agents are the people and institutions that build and deploy it.

Why This Moment, and Why the Vatican

There is a version of this argument that a secular body could make more precisely. The Vatican cannot, by its institutional nature, make it. And yet the encyclical may land precisely because it comes from a body that holds no interest in the outcome of the technology sector's competitive positioning. The Church has no stake in which firm leads AI development. It cannot be captured by Semiconductor Advisory Council. That distance may be the document's most valuable feature.

The DW News departure from X, reported on 25 May 2026, is a small data point in a larger pattern—the fracturing of the media environment along the same lines the encyclical identifies as structural. That departure was a commercial and reputational calculation made by a specific institution responding to a specific platform's choices. It was not a coordinated regulatory response. It is the kind of individual exit that satisfies no one and changes nothing systemically. The encyclical is, at its core, an argument against exactly that mode of response.

Pope Leo's document will be read by much of the Western press as a technology warning, parsed for its specific recommendations on AI development, assessed for its implications for the semiconductor sector, ignored by the systems it describes. That misreading is, in one sense, predictable: media outlets that depend on attention are poorly positioned to cover arguments about the cognitive architecture of the attention economy. But the document deserves a more careful reading than that—the more so because the problem it identifies is real, and the solutions it implies are harder than the ones the dominant conversation prefers.

The encyclical is not about AI. It is about what AI is allowed to do, and to whom, and without what kind of answer. That is a political question. The technology has been given, for now, a pass on the answer. Pope Leo is suggesting that pass is not indefinitely renewable.

This piece was filed from Vatican City. Monexus covered the encyclical's structural argument rather than its immediate technology-coverage angle, which most wire services led with.