Pope Leo XIV's first encyclical, Magnifica Humanitas, became an AI policy event on May 25, 2026. It is not a model release, but it belongs in the AI industry feed because it turns AI governance into a mainstream institutional priority and places Anthropic directly on the public stage.
Vatican News said the document was released on May 25, 2026, after being signed on May 15, the 135th anniversary of Pope Leo XIII's Rerum Novarum. The Vatican framed the new text as a response to the artificial intelligence era in the same way Rerum Novarum responded to the industrial revolution.
What happened on May 25
At the Vatican presentation on May 25, Pope Leo XIV described the AI shift as an epochal turning point. Vatican News reported that he called for AI to be disarmed, meaning freed from uses that turn it into an instrument of domination, exclusion or death.
The launch had an unusual AI-industry signal: Christopher Olah, a co-founder of Anthropic and a researcher known for interpretability work, was listed by Vatican News as one of the speakers at the presentation.
The policy argument
The encyclical's AI chapter focuses on responsibility and accountability at every stage of development. Vatican News summarized the document as calling for adequate AI policies and legal frameworks, independent oversight and user education.
The sharpest governance claim is that a more moral AI is not enough if the morality is determined by a few. That is a direct challenge to a market structure in which the largest model labs, cloud providers and platform companies control the resources, deployment channels and policy defaults for powerful systems.
War, work and data power
The Vatican summary puts autonomous weapons near the center of the argument. It says Pope Leo warned that AI can make conflict faster and more impersonal, lowering the threshold for violence by turning defense into threat prediction and victims into data.
The same framework extends to labor and platforms. Vatican News says the encyclical warns that AI systems can force workers to adapt to machine speed rather than being designed around human dignity. It also criticizes mass data collection, behavioral profiling and digital environments that turn personal lives into exploitable information.
The AP reported on May 25 that Pope Leo called for robust AI regulation and for developers to work for the common good rather than profit. That raises the political pressure around standards, liability, procurement rules and release governance.
Why Anthropic is part of the story
Anthropic did not publish the encyclical. The Vatican did. But Olah's presence matters because Anthropic has spent years positioning itself around interpretability, frontier-risk caution and constitutional AI.
That does not make Anthropic the owner of the policy agenda. It does make the event a reputational signal. When religious, civic and regulatory institutions look for AI expertise, they may increasingly choose labs that have invested in safety narratives, interpretability research and governance language.
ã.io read
Magnifica Humanitas is timely because it shows AI governance leaving the specialist lane. The same week the site covered OpenAI provenance and geometry reasoning, the Vatican put AI into one of its highest teaching formats and linked it to war, work, data, accountability and institutional power.
The industry takeaway is simple: AI policy is becoming moral infrastructure. Frontier labs will be judged not only by benchmark scores and product adoption, but by how their systems fit into public accountability, international governance and the human institutions that now see AI as a civilizational question.
ã.io publishes source-aware AI technology briefings for builders and operators.