At 11:30 AM Rome time today, Pope Leo XIV officially released his first encyclical, Magnifica Humanitas (Latin for 'Magnificent Humanity'), at the Vatican's Synod Hall. Sharing the stage with him was Christopher Olah, co-founder and head of interpretability research at Anthropic. The last time the Vatican did something like this was in 1891.
A Historic Joint Appearance
The podium featured six attendees: three cardinals (Víctor Manuel Fernández, Michael Czerny, Pietro Parolin), two female theologians (Anna Rowlands of Durham University and Léocadie Lushombo of Santa Clara University), and Anthropic's Christopher Olah. The Pope himself delivered the opening address and final blessing—a break from tradition, as encyclicals are usually presented by cardinals without the Pope present.
The encyclical was signed on May 15, exactly the 135th anniversary of Pope Leo XIII's Rerum Novarum, which addressed the impact of the Industrial Revolution on labor and is considered the foundation of modern Catholic social teaching. By choosing this date, the current Pope drew a clear parallel: AI in the 21st century is akin to industrialization in the 19th.
America Magazine quoted Charles C. Camosy: 'The head of the Catholic Church and one of the most consequential AI researchers and leaders in the world, on the same stage, presenting the same document. This has never happened before.'
What the Document Says
While the full text of Magnifica Humanitas is yet to be released, three key points have emerged from previews:
- AI cannot replace moral responsibility. The Pope has previously stressed that AI must not allow humans to shirk responsibility. In other words, when a machine makes a decision, a human must still sign off.
- Autonomous weapons are a red line. The Church continues its traditional 'just war' doctrine—removing 'meaningful human oversight' over life-and-death decisions crosses a red line. This directly targets lethal autonomous drones being developed by militaries worldwide.
- Efficiency is not the only standard. In fields like medicine, education, and art, there are 'goods internal to practices' that AI, if solely focused on efficiency, would crowd out. The Pope's message: you can use AI to speed up diagnosis, but the conversation between doctor and patient cannot be skipped.
Why Olah, Not Altman?
The Vatican invited Christopher Olah, not OpenAI's Sam Altman or DeepMind's Demis Hassabis. Olah leads interpretability research at Anthropic, focusing on understanding what large models are thinking, which neurons handle which functions, and what internal signals look like when a model lies. This is the most hardcore and least commercially rewarding branch of AI safety research.
Anthropic, from its founding, has positioned itself as prioritizing safety—a natural fit with the Church's message. The Vatican chose Olah not from the company with the most money, but the one with the most aligned approach.
Context: In February, the Trump administration targeted Anthropic for restricting military use of its models, leading Anthropic to sue the administration. The case is ongoing. Today's Vatican invitation is timely.
What This Means
An encyclical is the highest form of papal teaching. It guides the ethical judgments of 1.4 billion Catholics worldwide and influences policies at Catholic universities, hospitals, and schools.
Expected ripple effects:
- Global Catholic hospitals will re-examine ethical boundaries of AI diagnostic tools.
- Church-run schools will issue AI usage guidelines, possibly restricting student use of ChatGPT.
- Weapons manufacturers will face more opposition in Europe on autonomous weapons.
- AI company PR will repeatedly cite 'human in the loop' rhetoric.
Church pronouncements are not law, but the Church plays a long game. The 1891 Rerum Novarum influenced the policy frameworks of European social democratic parties throughout the 20th century—labor laws, social insurance, union rights all fell within its scope.
Whether Magnifica Humanitas will serve a similar anchoring role in the AI era is unknown. But by giving a seat to an AI company researcher, the Church has sent a signal: it intends to engage, not just observe.
Over the next six months, AI company PR departments are likely to be busy.
Sources: CocoLoop, Pope Leo XIV to launch his first encyclical, a document on artificial intelligence, with Anthropic's co-founder (PBS NewsHour); Pope Leo's encyclical comes just in time: AI is raising questions only religion can answer (America Magazine); Pope Leo will publish first encyclical, Magnifica Humanitas, on AI (America Magazine Vatican Dispatch)