Pope Leo XIV and Anthropic's Christopher Olah Call for AI to Be 'Disarmed' in Landmark Vatican Event

What the Encyclical Demands: 'Disarm' AI and Reject a 'Culture of Power'

Magnifica Humanitas is not a cautious reflection on technology; it is a sweeping manifesto with concrete governance demands. The document, signed on May 15 and presented with the pope himself delivering the final address, outlines a vision for AI that is explicitly at odds with the current trajectory of the industry .

1. Disarmament and 'Human-Friendly' AI
Pope Leo XIV directly called for AI to be "disarmed" and made "human-friendly," warning against "a race for ever more powerful algorithms and larger language models" that dehumanize people . This language frames the unbridled development of AI not as progress but as a form of arms racing that threatens human dignity itself.

2. Robust Regulation for the Common Good
The pope insisted that AI developers work for the common good rather than profit alone, calling for strong legal and regulatory frameworks to govern the technology . The encyclical was clear: the market alone cannot be trusted to guide AI development in a direction that serves humanity broadly.

3. Ban Autonomous Weapons
In one of its most concrete demands, Magnifica Humanitas condemned the use of AI in warfare, calling for prohibitions on lethal autonomous weapons systems. The pope stated it is morally impermissible to entrust irreversible, lethal decisions to AI systems .

4. Protect Workers
Echoing the Church's long tradition of social teaching on labor, the document warned that AI must not be used to exploit workers or replace human labor without just transitions and protections, pointing to "new forms of slavery" behind AI, from content moderators to miners .

5. Reject a 'Culture of Power'
Leo XIV denounced the techno-solutionist mindset that treats AI as an end in itself, comparing the attempt to build an AI future that excludes God to the biblical Tower of Babel . This theological frame positions unaccountable AI development as an act of hubris. The third chapter, titled "Technology and Dominance," develops this argument and calls for clear accountability throughout AI development .

The Vatican further strengthened its institutional commitment by establishing a new AI commission on May 16, tasked with coordinating AI-related activities across Vatican institutions and setting policies for AI use within the Holy See .

Christopher Olah's Remarks: Safety, Humility, and Moral Engagement

Complete verbatim transcripts of Olah's panel remarks have not been widely published, but coverage of the event and pre-event previews outline the main thrust of his message . Appearing as the lay speaker alongside top cardinals and theologians, Olah used the historic platform to push a vision of AI development that aligns with Anthropic's safety-first ethos.

Olah reportedly framed AI as a shared moral concern, arguing that AI development is not merely a technical challenge but a question of human dignity that demands broad societal input—including, explicitly, from religious traditions . He stressed the need for safety-first development, emphasizing that AI systems must be built with robust safeguards and that the industry must prioritize human wellbeing over unchecked competition . In a notable departure from techno-utopian rhetoric, Olah acknowledged that AI researchers do not have all the answers and must engage seriously with ethical frameworks outside the tech sector, pointing to Catholic social teaching as a valuable resource .

His presence was historic: it marked the first time a major AI executive shared the stage with a pope at the launch of a papal encyclical. The symbolism was not lost on observers, who noted that it suggests Anthropic sees the Church as a valuable institutional ally in advocating for safety norms in AI governance .

The Strategy Behind the Stage: Anthropic's Engagement with Religious Institutions

The May 25 event was not a spontaneous photo opportunity; it was the public culmination of a deliberate and long-running strategy by Anthropic to build bridges with religious and wisdom traditions.

In a May 19 press release, Anthropic revealed that it had spent "the past several months" organizing dialogues between its researchers and Vatican officials, including theologians and ethicists, to build shared understanding of AI's risks and moral dimensions . The company said its first round of discussions engaged "more than 15 religious and cross-cultural groups" as part of a broader effort to ground its AI development in moral traditions beyond the tech sector .

Anthropic's broader pattern is even more revealing. Months before the Vatican event, the company worked with a Catholic priest and other faith leaders to directly shape the "Claude Constitution," the set of guiding principles that govern how its AI behaves . Olah himself initiated the outreach, seeking help to encode ethical values into the company's systems. This reflects Anthropic's founding mission to build AI that is "helpful, honest, and harmless" and to conduct pre-deployment safety research—values that resonate naturally with the Vatican's emphasis on human dignity and caution .

Olah's attendance in Rome signals that the company views the Church as a serious institutional partner in the global conversation about AI regulation—one that brings moral authority and a global reach that no tech consortium or national regulator can match . The Vatican's decision to invite Olah to co-present gave that strategy its most significant endorsement yet.

In a world where AI development is accelerating, Magnifica Humanitas and its historic launch represent a clear demand: technology must serve, not dominate. And for Anthropic, standing with the Pope is a clear signal of which side of that debate it wants to be on.