The Vatican just cited Anthropic's interpretability research in a papal letter warning that AI could become "an instrument of domination, exclusion and death."
The Summary
- Pope Leo XIV released his first encyclical, "Magnifica Humanitas," calling for AI to serve humanity rather than concentrate power
- The document was signed on the 135th anniversary of "Rerum Novarum," the Church's foundational text on the Industrial Revolution
- Anthropic was consulted in the drafting process, signaling the Vatican is talking to frontier AI labs, not just ethicists
- The parallel to the first Industrial Revolution is deliberate: the Church is positioning AI as an economic transformation that requires moral guardrails
The Signal
The Vatican doesn't issue encyclicals lightly. These are the Church's heavyweight intellectual documents, meant to guide Catholics (and influence everyone else) for generations. Pope Leo XIV chose to make his first encyclical about AI, and he chose to release it exactly 135 years after Pope Leo XIII's "Rerum Novarum," the 1891 letter that established Catholic social teaching on labor rights, capitalism, and the dignity of workers in the machine age.
The timing is not subtle. Leo XIII wrote "Rerum Novarum" when factories were replacing craftsmen, when capital was concentrating, when entire classes of people were being rendered economically obsolete. Leo XIV is drawing a direct line to today: AI is the new factory, the new steam engine, the new force that will either liberate people or crush them.
"AI could become an instrument of domination, exclusion and death."
What's unusual here is who the Vatican invited to the table. Anthropic was consulted during the drafting of "Magnifica Humanitas". That's not a token gesture. That's the Church saying it needs to understand the actual technology, not just the philosophy. It suggests the Vatican believes interpretability research, the kind Anthropic publishes on how models actually work inside, matters for moral questions about AI deployment.
This is different from the usual corporate ethics-washing. When a company funds a university AI ethics center, everyone knows the score. When the Pope cites your research in a document meant to shape global moral consensus, that's a different kind of attention. It raises questions:
- Did Anthropic shape the encyclical's technical framing?
- Is the Vatican endorsing certain AI safety approaches over others?
- What does it mean for a frontier lab to have theological credibility before it has widespread commercial deployment?
The encyclical's core argument, that AI must serve humanity rather than powerful elites, lands differently when you know Anthropic was in the room. It's not an abstract call for justice. It's a warning from people who understand what's technically possible about what's ethically permissible.
The Implication
This encyclical will be cited in AI policy debates for the next decade. Lawmakers in Catholic-majority countries now have Vatican backing to regulate AI on moral grounds, not just economic or security ones. That shifts the conversation from "can we build this" to "should we deploy this, and for whose benefit."
For AI companies, this is a new kind of pressure. You can ignore activists. You can lobby against regulators. But when 1.4 billion Catholics are told by their Pope that your technology could be "an instrument of domination," you're playing a different game. Watch for how frontier labs respond. If they lean into interpretability and alignment research as moral imperatives, not just technical ones, you'll know the Vatican's framing landed.