Pope Leo XIV's First Encyclical Takes on AI's Human Cost

1891 and 2026: Two Encyclicals Signed on the Same Date

On May 15, 1891, Pope Leo XIII issued the encyclical¹ Rerum Novarum (“On New Things”), addressing the treatment and rights of workers. It was a document that asked, as industrialization drove production ever higher, under what conditions workers actually lived and labored.

135 years later, on May 15, 2026, Leo XIV signed his first encyclical, Magnifica Humanitas. The Vatican announced that the document addresses the protection of humanity in the age of AI, with the full text to be released on May 25. Alongside the pope, cardinals, and theologians, the release event is set to include Anthropic co-founder Christopher Olah.

I found myself curious about how the Catholic Church’s long history of engaging with labor issues connects to this new discussion of AI. And I was struck, too, by the choice to invite a technology researcher to sit alongside the church officials at the announcement.

Worker Rights in the 1891 Encyclical

In the industrialization of the 19th century, long working hours, child labor, and poverty were pressing social problems. Debate raged over who should capture the gains from rising production, and who bore responsibility for protecting workers’ health and livelihoods.

Rerum Novarum affirmed private property while criticizing the treatment of workers as mere instruments of production. Its position was that economic activity, too, carries an obligation to safeguard human dignity and welfare.

The document addressed a living wage, rest, protection from excessive labor, and the role of workers’ associations. It went on to become a foundational reference in the development of Catholic social teaching.

A papal encyclical doesn’t automatically become law in any country. But it can serve as a teaching that believers and Catholic institutions draw on when weighing social issues. I think the influence of a document like this should be measured not by the moment of its release, but by the debate and practice it sets in motion afterward.

Why Leo XIV Connected AI to Labor

AI’s advance raises questions about how we work and how the gains get shared.

Large language models² are already being used for office work like writing, analysis, and coding. Alongside their potential to boost productivity, we need to look at how jobs are changing, where accountability lies, and how influence is concentrating in a handful of companies. The technological conditions are different from the Industrial Revolution, but the question of who captures the benefits and who bears the costs of change is one we can still connect to that same thread.

Leo XIV’s given name is Robert Prevost. He earned a bachelor’s degree in mathematics from Villanova University in 1977 and spent many years doing pastoral work in Peru. His background in mathematics is intriguing, but it alone doesn’t explain his stance on AI.

The reason he himself gave is closer to Catholic social teaching. In May 2025, addressing the College of Cardinals to explain his choice of papal name, he cited Leo XIII’s response to the labor question. He said that today, too, AI and new industrial change pose challenges to human dignity, justice, and labor.

The Vatican’s advance notice specified that the new encyclical’s signing date falls on the 135th anniversary of the publication of “Rerum Novarum.” This officially underscores the link between the two documents. Still, what concrete standards the new encyclical will actually contain has to wait for the full text’s release on May 25.

Cardinal Víctor Manuel Fernández, Prefect of the Dicastery for the Doctrine of the Faith, and Cardinal Michael Czerny, Prefect of the Dicastery for Promoting Integral Human Development, are set to attend the launch event. Professor Anna Rowlands of Durham University, Professor Leocadie Lushombo of Santa Clara University, and Christopher Olado have also been announced as speakers.

Following closing remarks from Cardinal Secretary of State Pietro Parolin, the Pope’s address and blessing are scheduled. What stands out is the lineup itself — church leaders, theologians, and technology researchers sharing the same stage.

A Tech Researcher Joins the Announcement

Olah is a researcher working on AI interpretability³ — the field trying to understand what computations happen inside AI systems and what actually drives their outputs.

He’s done research at Google Brain and OpenAI, and joined Anthropic as a co-founder in 2021. The Vatican’s announcement of the event lists him with the co-founder title and notes that he leads interpretability research.

This kind of research connects directly to questions like: how far can we actually explain an AI’s judgments, and how do we hold anyone accountable for decisions that affect people? That’s the significance I see in having a tech researcher join this event.

The Vatican has engaged with tech companies on AI ethics before. In 2020, the Pontifical Academy for Life launched the “Rome Call for AI Ethics”⁴ together with Microsoft, IBM, and others. It laid out six principles covering transparency, inclusion, responsibility, fairness, reliability, and security and privacy.

Anthropic has also recently clashed with the U.S. government over the scope of military use of AI. The company has argued that large-scale domestic surveillance of Americans and fully autonomous weapons use should be restricted. A February 27 company statement lays out that position.

Litigation has since followed, touching on things like the U.S. government’s supply-chain risk designations. One of the questions at stake is how far companies can go in setting conditions on the use of their own AI.

Still, the Vatican hasn’t framed Olah’s invitation as an endorsement of either side in this dispute. The presence of a single co-founder isn’t enough to conclude that the Church and the company align on every issue. What’s officially confirmed is simply that he’s taking part in the encyclical’s unveiling.

Oswarld’s Lens

Watching this announcement, I was reminded of something I used to grapple with when building go-to-market strategy.

When you introduce a new product or technology, the conversation doesn’t end with a rundown of specs. You also need criteria for what counts as a good outcome and what kinds of use are unacceptable. Who sets those criteria can shape everything that follows.

The Vatican, too, is trying to lay out criteria for judging AI, grounded in Catholic social teaching. That’s a different kind of thing from a corporate product spec, but I think there’s real significance in evaluation criteria being debated outside the companies that build the technology.

An encyclical isn’t a document that directly regulates companies the way a national AI law would. And the mere existence of an encyclical doesn’t automatically change how society behaves. What matters is how educational institutions, researchers, companies, and the faithful actually receive and apply its content.

One thing I learned from building corporate strategy is that who is explaining something, and in what setting, matters as much as what is being said. This time, an AI researcher stands alongside the Pope and theologians. I’ll be watching to see whether this lineup helps different fields ask each other better questions.

Closing

When the full text is released on May 25, I want to read exactly what evidence it uses to explain AI’s impact on labor and human judgment.

I’m also curious whether it lays out concrete criteria for issues already under debate, like autonomous weapons, surveillance, and worker protections. What matters more than the list of attendees is checking, firsthand, the evidence and proposals the document actually puts forward.

Looking good—no Hangul, all links/images/footnotes match. Output unchanged:

References & Further Reading

Primary sources

• “Pope Leo XIV’s first encyclical Magnifica humanitas to be published May 25”, Vatican News, 2026.05.18. : This is the original text of the Vatican’s official announcement.
• “Pope Leo to present his encyclical on AI alongside Anthropic co-founder”, National Catholic Reporter, 2026.05.19. : This report covers who attended the presentation event and how it was structured.
• “Pope Leo’s encyclical comes just in time: AI is raising questions only religion can answer”, America Magazine, 2026.05.19. : A column that directly compares Anthropic’s AI constitution with the Catholic concept of human dignity. If you’re interested in this topic, I’d recommend starting here.

Background

• Pope Leo XIII, “Rerum Novarum”, 1891. : The original text of the 1891 encyclical addressing labor, private property, wages, and workers’ associations.
• “Rome Call for AI Ethics”, Pontifical Academy for Life, 2020.02.28. : The starting point of the Vatican’s AI ethics framework. Its six principles are laid out in just two pages.
• “Anthropic sues the Trump administration over ‘supply chain risk’ label”, NPR, 2026.03.09. : An article covering the dispute between Anthropic and the US government.
• “Pope Leo sets Catholics on collision course with AI”, Axios, 2026.05.14. : A concise summary of the encyclical’s expected contents and the arc of AI discussions dating back to Pope Francis’s papacy.

The author is Oswarld (Kwangseob Ahn). Current roles: Adjunct Professor at Sejong University, Strategy Consultant at INLEVEL9. Career, research, books, and recent work are kept current on the About page. Latest · July 2026: HEMA-2: A Consolidation-Aware Tri-Memory Architecture with Multi-Channel Scheduling for Lifelong Conversational AI.

Footnotes

1. Encyclical: A type of official document through which the Pope conveys teaching on matters of faith, morals, and social issues. The form of the document alone doesn’t mean every sentence carries the force of an infallible declaration or civil law. ↩

2. Large Language Model (LLM): The underlying technology behind AI services like ChatGPT and Claude. It refers to AI models trained on vast amounts of text data to read and write like a human. ↩

3. Interpretability: A research field aimed at explaining an AI’s internal computations and behavior in ways humans can understand. Current methods can’t fully explain every step of the decision-making process. ↩

4. Rome Call for AI Ethics: An AI ethics declaration launched in 2020 by the Pontifical Academy for Life together with Microsoft, IBM, and others. It sets out six principles: transparency, inclusion, accountability, impartiality, reliability, and security and privacy. ↩