Pope Leo XIV's AI Encyclical 'Magnifica Humanitas' Drops May 25 — With Anthropic's Co-Founder on Stage

On Monday, May 25, in the Vatican’s Synod Hall, Pope Leo XIV will publish the Catholic Church’s most direct engagement with the age of artificial intelligence. The document is called Magnifica Humanitas — “Magnificent Humanity” in Latin — and it is the first papal encyclical in history to address AI as its central subject.

Standing at the podium alongside the Pope will be Christopher Olah, co-founder of Anthropic and one of the most influential researchers in AI interpretability. The pairing of the world’s most visible religious leader with one of Silicon Valley’s most prominent AI safety researchers is not accidental. It announces, plainly, that the church and the technology industry are now in direct conversation about the same set of questions.

The Historical Echo

The timing of the document’s signing carries deliberate weight. Pope Leo XIV signed Magnifica Humanitas on May 15, 2026 — exactly 135 years after Pope Leo XIII signed Rerum Novarum on May 15, 1891. Rerum Novarum was the church’s foundational response to the industrial revolution: a defense of workers’ rights, a critique of unbridled capitalism, and an assertion that labor has inherent dignity that markets cannot fully price.

The symmetry is intentional. The Vatican is framing AI not as a technological novelty but as a civilizational shift comparable to industrialization — one that demands the same quality of moral response that the church offered to the upheaval of the 19th century.

Leo XIV, born Robert Francis Prevost, was elected Pope in May 2025 after the death of Pope Francis. His papacy has been characterized by a strong engagement with contemporary social questions, and he signaled in his first months that AI would be a central preoccupation of his tenure.

What the Encyclical Covers

Based on pre-publication summaries circulated to the Vatican press corps and reporting from America Magazine and the National Catholic Reporter, Magnifica Humanitas addresses three broad domains.

The first is the dignity of the human person in an age of automated decision-making. The encyclical is expected to argue that algorithmic systems used in criminal justice, employment screening, credit, and healthcare create a new form of structural injustice when they make consequential decisions about individuals without meaningful transparency or accountability. The document draws on the natural law tradition to assert that human beings have a right to be judged by other human beings — or, at minimum, by systems that can be held accountable in human terms.

The second domain is labor. Magnifica Humanitas reportedly extends the analysis of Rerum Novarum and its successors directly to AI-driven automation. The encyclical is expected to defend workers’ rights to retraining, to fair transition support when AI displaces their roles, and to meaningful economic participation in the productivity gains that automation generates. This is not a call to stop AI development — sources familiar with the document describe it as explicitly pro-technology — but a call to distribute its benefits justly.

The third domain is what the Vatican is calling “epistemic rights” — the right of individuals not to be manipulated by AI systems optimized for engagement, addiction, or the shaping of political opinion. The encyclical engages directly with the mechanics of recommendation systems and generative AI, arguing that systems designed to maximize time-on-platform at the expense of human flourishing represent a moral harm, not merely a design choice.

Why Christopher Olah

The decision to include Christopher Olah in the encyclical’s public presentation is striking on multiple levels. Olah is one of the original architects of AI interpretability research — the field dedicated to understanding how neural networks actually produce their outputs, rather than treating them as black boxes. He is also a practicing Christian, and has spoken publicly about the relationship between his faith and his conviction that AI safety is a moral imperative.

His presence signals that the Vatican engaged seriously with technical AI research in drafting the document, not merely with philosophical abstractions. It also suggests that Anthropic — which has positioned itself as the AI company most explicitly focused on safety — sees value in associating itself with the church’s moral framework at a moment when AI regulation is contested everywhere.

The combination is notable given the political context. Anthropic has recently been at the center of controversy over the potential use of its models in US military and surveillance applications, with reports of pressure from the Trump administration to loosen the company’s usage restrictions. Aligning with a major international moral authority may not be purely coincidental timing.

The Broader Reception

The encyclical has already attracted attention beyond Catholic circles. JD Vance — himself a Catholic convert — said he was “looking forward to reading it.” European AI regulators have indicated they will study the document. Several AI companies have signaled they will respond formally to the church’s arguments.

Bishop Daniel Flores, in an essay published ahead of the release, made the argument that reading the encyclical requires genuine engagement — “Don’t let ChatGPT read it for you,” he wrote, an injunction that became an immediate headline.

Magnifica Humanitas will not carry the force of law. Encyclicals are not legislative documents. But the Catholic Church has 1.4 billion members and an institutional presence in more countries than any government. When it speaks clearly on a moral question, it shifts the frame of public debate — sometimes slowly, sometimes with remarkable speed.

The AI industry has spent years debating its ethical obligations largely within a bubble of technologists, ethicists, and regulators. On May 25, one of the oldest moral institutions in human civilization enters that debate with a document it spent years preparing. The conversation is about to get considerably more complicated.