A new essay from Reason Magazine has drawn an unlikely comparison: Anthropic, the San Francisco AI safety company behind Claude, and the Vatican. The piece argues that both institutions are engaged in a battle for moral authority over artificial intelligence, deploying the language of existential risk and human dignity to steer how society thinks about the technology. The comparison is provocative, but it surfaces a real and underreported tension in how different power centers are racing to set the ethical terms for AI's future.
Two Frameworks, One High-Stakes Question
The Reason analysis centers on a simple observation: Anthropic and the Catholic Church are both trying to answer the same question, namely, what does AI mean for humanity, and who gets to decide? Anthropic approaches this through its Constitutional AI methodology, published safety research, and a stated mission to develop AI for the long-term benefit of humanity. The Vatican, meanwhile, has issued its own guidance through documents like the "Rome Call for AI Ethics" and Pope Francis's repeated interventions on algorithmic bias, surveillance, and the commodification of human attention. Both are moral frameworks. Neither is purely technical.
Key Facts
- Reason Magazine frames the Anthropic-Vatican comparison as a contest for moral authority over AI governance.
- The Vatican's "Rome Call for AI Ethics" has been signed by major tech firms and international organizations.
- Anthropic's Constitutional AI trains models using a set of principles, some of which overlap with religious ethics on human dignity.
- Pope Francis addressed AI directly at the G7 summit in Italy in 2024, the first pope to attend such a meeting.
- Libertarian outlets like Reason often critique both religious and corporate paternalism in tech policy debates.
Where the two diverge is significant. Anthropic's ethics are consequentialist and empirical, built around measurable outcomes like harmlessness and helpfulness. The Church's framework is rooted in natural law and the sanctity of the human person, a basis that leads to very different conclusions on topics like AI-generated content, autonomous weapons, and the replacement of human labor. Reason's piece, true to its libertarian editorial line, treats both as varieties of top-down moralizing that risk crowding out individual choice and open markets. That reading is contestable, but it captures a genuine anxiety about who writes the rules.
"The real question isn't whether AI should have ethics. It's whose ethics, enforced how, and with what accountability."Reason Magazine
Why This Debate Is Getting Louder
The timing of the article is not accidental. Governments and international bodies are moving toward binding AI regulation, and every major actor with a stake in the outcome is trying to shape the conversation. Earlier this year, Anthropic's CEO joined peers from OpenAI and Google at the G7, where AI governance was a central agenda item. Pope Francis had already made history by attending the 2024 G7 in Italy to speak on AI. The overlap between these two events illustrates how thoroughly the debate has expanded beyond engineering teams and into the domain of politics, theology, and philosophy.
Anthropic's position in this landscape is complex. The company occupies an unusual space in the AI model market, simultaneously building powerful frontier systems and advocating loudly for caution and oversight. Critics, including some in the libertarian press, argue that safety-focused companies benefit from regulatory frameworks that disadvantage smaller competitors. Supporters counter that Anthropic's research into alignment and interpretability is genuine and necessary. Neither characterization is complete on its own.
The Reason essay does not resolve these tensions, nor does it try to. Its value lies in forcing a question that technical communities sometimes avoid: when an AI company publishes a set of values and trains a model on them, it is making a moral choice. That choice deserves scrutiny from many directions, not only from regulators or ethicists, but from religious traditions, civil libertarians, and the public at large. Whether Anthropic and the Pope are rivals or accidental allies in that project depends entirely on which part of the agenda you are examining.
For now, the debate is likely to intensify as AI capabilities expand and governance frameworks harden. The question of who holds moral authority over these systems will not be settled by any single essay, summit, or papal encyclical. It will be worked out slowly, in public, across a very wide set of institutions that are only beginning to understand what they are actually arguing about.
