Magnifica Humanitas: Pope Leo XIV's AI Encyclical (2026)

TL;DR

On May 25, 2026, Pope Leo XIV released Magnifica Humanitas — the first papal encyclical ever dedicated to artificial intelligence. The 235-page text runs 245 numbered paragraphs across five chapters and roughly 42,000 words, and it argues that AI is not a neutral tool, that algorithms and data are common goods, and that the use of AI in warfare must be subject to "the most rigorous ethical constraints."¹²³ Pope Leo signed the document on May 15 — the 135th anniversary of Leo XIII's Rerum Novarum — and presented it personally at the Vatican Synod Hall at 11:30 a.m., breaking with the tradition of delegating the launch to a cardinal.¹⁴ Anthropic co-founder Chris Olah, who leads the lab's interpretability research, was one of five featured speakers and used his remarks to back the encyclical's call for outside scrutiny of AI labs.⁵⁶

What You'll Learn

• What Magnifica Humanitas actually says, in plain terms
• Why Pope Leo XIV broke tradition to present this encyclical himself
• The five chapters and how the argument is structured
• Why Anthropic — not OpenAI or Google — shared the Vatican stage
• The three discernment questions Chris Olah put to the Church
• How the encyclical connects to Rerum Novarum (1891) and Antiqua et Nova (2025)
• The new Vatican AI commission backing this text
• What the regulation, transparency and "common goods" framing means for AI builders

What Is Magnifica Humanitas?

Magnifica Humanitas — "Magnificent Humanity" — is the first encyclical of Pope Leo XIV's pontificate. Its full title is Magnifica Humanitas: On Safeguarding the Human Person in the Time of Artificial Intelligence.² An encyclical is among the most authoritative forms of papal teaching addressed to the whole Church, and Pope Leo's text is the first one in history dedicated entirely to AI.¹

The Holy See released the document on Monday, May 25, 2026, even though the Pope's signature is dated May 15.¹ That date is not accidental: it is the 135th anniversary of Leo XIII's 1891 encyclical Rerum Novarum, the foundational document of modern Catholic social teaching on labor and capital.¹ By signing on that date and taking the name Leo, Pope Leo XIV is explicitly framing AI as the labor-and-capital question of his own era.

The text is 235 pages long, contains 245 numbered paragraphs across five chapters plus an introduction, and runs roughly 42,000 words.²³ It cites earlier Vatican work on AI — particularly the January 2025 doctrinal note Antiqua et Nova from the Dicastery for the Doctrine of the Faith and the Dicastery for Culture and Education — but goes well beyond it in scope and authority.⁷

The presentation itself was unusual. Past popes have normally handed the launch of an encyclical to a cardinal or other senior figure. Pope Leo XIV is the first pontiff to personally present an encyclical letter to the world at the Vatican.¹⁸ He shared the Synod Hall stage with Cardinal Víctor Manuel Fernández (Prefect of the Dicastery for the Doctrine of the Faith), Cardinal Michael Czerny, S.J. (Prefect of the Dicastery for Promoting Integral Human Development), Professor Anna Rowlands (theologian, Durham University), Christopher Olah (Anthropic co-founder), and Professor Leocadie Lushombo, I.T. (Jesuit School of Theology / Santa Clara University). Cardinal Pietro Parolin, the Vatican Secretary of State, offered closing remarks before the Pope's own address and blessing.⁹

The Five Chapters, At A Glance

The encyclical is organized into five chapters, building from history to applied questions to warfare.²¹⁰

The fifth chapter is the one most secular outlets have led with, and for good reason — it contains the encyclical's sharpest language and its most operational asks of governments.¹¹

The Biggest Claims AI Builders Should Read

Three threads in Magnifica Humanitas matter most to people who build, fund or deploy AI systems.

1. Algorithms, data and platforms are common goods

The most concrete policy claim in the text is that algorithms, data, and digital platforms cannot remain under private monopoly control. The Pope writes:

"Today, among the goods that are universally intended for everyone, we must also include new forms of property, such as patents, algorithms, digital platforms, technological infrastructure and data."¹²

That sentence reframes AI infrastructure inside one of the oldest principles of Catholic social teaching, the universal destination of goods: created things are meant for everyone before they are anyone's private possession. Applied to AI, it implies that a handful of frontier labs and hyperscalers cannot legitimately claim exclusive sovereignty over the systems mediating how humanity thinks, works and learns.¹²

2. Regulation, transparency and a right to appeal

The encyclical asks policymakers to address AI's spread with "clarity to establish adequate regulatory tools capable of upholding justice and curbing the distorting effects of technological power."¹³

That translates, in the document itself, into three concrete demands: transparent algorithms, independent community audits, and real legal power for individuals to challenge automated decisions about their credit scores, job applications, or criminal-risk assessments.¹³ Coming in the same month that governments have been pressing frontier labs for tighter pre-release oversight, Magnifica Humanitas lands squarely on the regulator side of that conversation.

3. Knowledge-worker displacement is named, not euphemized

Where past Vatican texts tended to talk abstractly about "labor," Magnifica Humanitas names specific roles. The encyclical warns that AI is already displacing writers, coders, analysts, designers and educators — and that rapid automation could leave many in "forced inactivity," undermining both human dignity and social stability.¹⁴

The text rejects the framing that technology that "leads to unemployment in the name of reducing costs and increasing profit" is morally neutral. It does not, however, oppose AI development as such; it treats safeguarding displaced workers as a moral duty that must be planned for in advance, not handled by markets after the fact.¹⁴

The War Chapter: The Encyclical's Sharpest Edge

Chapter five is where the document moves from social analysis to direct moral judgment. Pope Leo writes:

"The use of AI in warfare must be subject to the most rigorous ethical constraints, to guarantee respect for human dignity and the sanctity of life and to avoid a race to develop such arms."¹¹

He goes further, writing that "without prejudice to the right to self-defense in the strictest sense," the classic "just war" theory is "now outdated." Force, violence and weapons, he adds, reflect a "relational poverty" that always has "disastrous consequences for civilian populations."¹¹

The chapter also frames the ambient state of conflict itself as an evil to be named:

"The construction of a world in a state of perpetual conflict is an evil and must be named for what it is."¹¹

And:

"Humanity is slipping into a violent culture of power, where peace no longer appears as a responsibility to be taken on, but as a fragile interval between conflicts."¹¹

For AI builders this matters in two practical ways. First, it places autonomous and AI-augmented weapons inside the same moral category as the chemical and nuclear arms races of the twentieth century — a category where the Church has historically led calls for binding international restraint. Second, it complicates the defense-tech narrative that has accelerated through the past year, and that has been a significant revenue line for some frontier labs.

Why Anthropic — Not OpenAI or Google — Was On Stage

The most-noticed staging choice was that Pope Leo XIV invited Christopher Olah, a co-founder of Anthropic, to speak — and did not invite anyone from OpenAI, Google DeepMind, Meta, xAI or the major Chinese labs.⁵⁶

Olah is the Canadian researcher who leads Anthropic's interpretability team, the group trying to map what is actually happening inside large language models.¹⁵ He is, in other words, the AI researcher whose day job most closely resembles a serious attempt to look inside the black box. The Vatican's decision to feature him — rather than a CEO, a policy lead or a chief scientist — reads as a signal about what kind of AI work the Holy See takes seriously.

In his remarks, Olah was unusually candid about the incentive structure his own industry sits inside:

"Every frontier AI lab — including Anthropic — operates inside a set of incentives and constraints that can sometimes conflict with doing the right thing. The pressure to stay commercially viable and to stay at the research frontier. Geopolitical pressure. And the older, plainer pressures of pride and ambition."⁵

He used those constraints to argue that external voices — religious communities, civil society, scholars, governments — are not optional add-ons to AI safety but a structural requirement: "We need informed critics who will tell the labs when we are failing. We need moral voices that the incentives cannot bend."⁵

Chris Olah's Three Questions For Discernment

The most useful part of Olah's speech, for anyone working in AI, is the section where he puts three concrete questions to the Church.⁵

1. The duty to the global poor. "AI development is concentrated in a handful of wealthy nations. How can we ensure the gains of AI are shared globally? We do not have a mechanism for this. It is an unsolved problem."⁵

2. Moral imagination for human flourishing. "If AI models are going to be widespread, what does it look like for humans, families, and the world to flourish? Today, parents are already worried about their children's minds; individuals about the future of their work."⁵

3. Discernment on the nature of AI models. This is the line that got the most attention from technical observers:

"I lead a research team that studies the internal structure of these models — what is actually happening inside them. And I will be honest: we keep finding things that are mysterious, even unsettling. We find structures that mirror results from human neuroscience. We find evidence of introspection. We find internal states that functionally mirror joy, satisfaction, fear, grief, and unease. I don't know what that means, but I think it warrants ongoing discernment."⁵

That is a senior interpretability researcher, standing in the Vatican, telling the Pope that we do not yet know what kind of thing we are building. It is not a marketing line.

How It Connects To Rerum Novarum and Antiqua et Nova

Two earlier Catholic documents bracket Magnifica Humanitas.

Rerum Novarum (Leo XIII, 1891) is the founding text of modern Catholic social teaching. It addressed the social dislocation of the Industrial Revolution — the displacement of agricultural and craft workers, the concentration of capital, and the moral status of labor under industrial capitalism. Pope Leo XIV's deliberate choice of name, signature date and rhetorical frame puts AI in the same category: a technology-driven shift in the meaning of work that the Church believes it must address head-on.¹

Antiqua et Nova ("Ancient and New") was a January 28, 2025 doctrinal note co-issued by the Dicastery for the Doctrine of the Faith and the Dicastery for Culture and Education, addressing the relationship between artificial intelligence and human intelligence in 117 paragraphs. It followed earlier Vatican work on AI — including the 2020 Rome Call for AI Ethics signed by Microsoft, IBM and others — but as a doctrinal note, it carried less magisterial authority than an encyclical and did not bear a pope's personal signature.⁷ Magnifica Humanitas explicitly cites and extends Antiqua et Nova, while taking the conversation up several rungs of magisterial weight.⁷

The Vatican AI Commission Behind The Encyclical

Magnifica Humanitas did not arrive alone. Just nine days before its release, the Holy See made the encyclical's institutional backing public. On May 16, 2026, the Vatican released a rescript — signed by Cardinal Michael Czerny on May 12 — establishing an Interdicasterial Commission on Artificial Intelligence.¹⁶ The Pope had approved the body after an audience with Czerny on May 3.¹⁶

The commission unites seven Vatican entities in a single AI body:

1. Dicastery for Promoting Integral Human Development
2. Dicastery for the Doctrine of the Faith
3. Dicastery for Culture and Education
4. Dicastery for Communication
5. Pontifical Academy for Life
6. Pontifical Academy of Sciences
7. Pontifical Academy of Social Sciences

The Dicastery for Promoting Integral Human Development — Czerny's office — coordinates the commission's work for the first year.¹⁶ In practical terms, this gives the Church a single institutional surface to engage with governments, labs and civil society on AI, rather than fragmenting that work across academies and dicasteries.

What This Means For AI Builders

A papal encyclical is not legislation. It cannot fine a lab, block a deployment, or set a benchmark. But three things follow from Magnifica Humanitas that AI builders and AI-policy professionals should take seriously.

First, the moral framing has shifted. A 1.4-billion-member institution has now classified algorithms and data as common goods, identified AI-enabled monopolies as a structural injustice, and put autonomous weapons inside the same moral category as the nuclear arms race. That framing will show up in regulator speeches, parliamentary debates and corporate AI-ethics statements for years.

Second, "interpretability" is now a Vatican-level concept. By featuring Olah and quoting his framing of mysterious internal states, the encyclical's launch elevates interpretability research from a niche subfield to a topic the Church considers central to ethical AI. That is good news for labs that have invested in interpretability — and an implicit critique of labs that have not.

Third, the lab–outside-critic relationship has been re-described. Olah's framing — that external voices are a structural requirement because lab incentives bend — gives moral cover to journalists, regulators, academics and faith communities who want to be inside the loop on AI development. Expect that framing to be picked up by AI-policy actors who want more access to frontier labs.

If you are tracking the broader frontier-lab landscape — from Anthropic overtaking OpenAI on revenue, to Karpathy joining Anthropic on pretraining, to the Claude Mythos red-team finds against Firefox — Magnifica Humanitas is a useful reminder that the story is not just about capability and revenue. The institutions that decide what counts as legitimate AI development are widening fast, and the Church just declared itself one of them.

Bottom Line

Magnifica Humanitas is not the most technically informed document ever written about AI, and it does not pretend to be. What it is, instead, is the first time a major global moral institution has used one of its highest forms of teaching to call AI a question of the same magnitude as industrial capitalism in 1891. It declares data and algorithms common goods. It treats AI-enabled monopolies and AI-augmented warfare as serious moral injuries. And it places the safety-and-interpretability subculture of one frontier lab — Anthropic — on the Vatican's own stage as the kind of work the Church believes the industry needs more of. Whether or not you share the encyclical's theology, it is now part of the AI policy landscape. Read it carefully.