Ethics
Magnifica Humanitas: Pope Leo XIV's 42,000-Word Reckoning With Artificial Intelligence
In the first encyclical of his papacy, Pope Leo XIV issued the most comprehensive moral analysis of artificial intelligence ever produced by a major religious institution. The 42,000-word document calls for AI to be 'disarmed' of its competitive logic — and names Anthropic by name.
By Patrick T ·
On 30 May 2026, Pope Leo XIV released Magnifica Humanitas — Latin for 'magnificent humanity' — the first encyclical of his papacy and the most detailed moral analysis of artificial intelligence ever produced by a major religious institution. At 42,000 words, the document is longer than most doctoral theses. It is also, in the context of the AI policy debate, something genuinely unusual: a sustained philosophical argument, grounded in centuries of Catholic social teaching, that takes the technology seriously enough to engage with it on its own terms.
The encyclical arrives at a moment when the Vatican's relationship with the AI industry has become unexpectedly close. Anthropic, the AI safety company behind Claude, has been in active dialogue with the Vatican's ethics office for several years. The encyclical names Anthropic explicitly — a striking gesture that signals the document is not a generic warning from an institution unfamiliar with the technology, but a considered response from one that has been watching from the inside.
"To disarm does not mean rejecting technology, but preventing it from dominating humanity." — Pope Leo XIV, Magnifica Humanitas
The central argument of Magnifica Humanitas is built around a single metaphor: AI has been 'armed' — shaped by the logic of geopolitical and commercial competition — and must be disarmed. The Pope is not calling for a ban. He is calling for a change in the conditions under which AI is developed and deployed. 'Technology is never neutral,' he writes, 'because it takes on the characteristics of those who devise, finance, regulate and use it.' AI developers, therefore, 'bear a particular ethical and spiritual responsibility, for every design choice reflects a vision of humanity.'
The Subsidiarity Argument
One of the encyclical's most substantive contributions to the AI governance debate is its invocation of subsidiarity — the Catholic social teaching principle that decisions should be made at the most local level possible. Leo applies this directly to AI development, arguing that communities and civil society organisations must not be 'reduced to passive recipients of decisions made elsewhere.' The concentration of AI capability in a small number of large technology companies, the document implies, is not merely an economic problem. It is a structural violation of the conditions under which human dignity can be maintained.
The encyclical also takes a clear position on data ownership: 'ownership of data cannot be left solely in private hands but must be appropriately regulated.' This is a more direct intervention into technology policy than most papal documents venture, and it aligns the Vatican's position with the European Union's data governance framework more closely than with the American model of largely unregulated data collection.
The Transhumanism Critique
The encyclical's most philosophically ambitious section takes on transhumanism and posthumanism — the intellectual traditions that treat human biological limitations as problems to be solved through technology. Leo's response is not a simple rejection. It is a reframing: 'humanity flourishes not despite limitations, but often through them.' The capacity for compassion, for generosity, for spiritual experience — these are not bugs in the human system to be patched by AI. They are, in Leo's account, the very features that make human life worth living.
'If the human being is treated as something to be perfected or surpassed,' the encyclical warns, 'it becomes easier to accept that some lives are less useful, less desirable or less worthy. In the name of progress, necessary sacrifices may begin to be justified, placing the burden on the most vulnerable in pursuit of a supposed optimization of the species.' The passage is a direct challenge to the effective altruist and longtermist currents that have shaped the thinking of several major AI labs.
The Anthropic Connection
The explicit naming of Anthropic in a papal encyclical is without precedent in the history of technology. It reflects a relationship that has been developing quietly since at least 2023, when the Vatican's Dicastery for Culture and Education began engaging with AI companies on questions of ethics and human dignity. A source engaged in the Vatican's tech outreach, speaking ahead of the encyclical's release, described tech leaders as 'interested in wisdom from the church' regarding 'how best to serve humanity.'
Anthropic's response to the encyclical has been measured. The company's interpretability team — which works on understanding what is happening inside AI models — has been cited in the document as an example of the kind of transparency work the Vatican considers necessary. This is a notable endorsement: the Pope is not simply criticising AI companies, but distinguishing between those doing safety work and those that are not.
What the Encyclical Does Not Do
Magnifica Humanitas does not propose specific regulatory mechanisms. It does not call for a moratorium on AI development. It does not endorse any particular governance framework. What it does is something that the secular AI policy debate has largely failed to do: it provides a coherent account of what is at stake for human beings — not just as workers or citizens or consumers, but as persons with an interior life that cannot be reduced to data.
Whether that argument will influence the engineers, investors, and policymakers who are actually shaping AI's trajectory is an open question. The encyclical is addressed, formally, to Catholics. But its audience, in practice, is anyone who has found the utilitarian calculus of the AI safety debate insufficient — anyone who suspects that the question of what AI is doing to human beings is not fully captured by benchmarks, alignment scores, or capability evaluations. For that audience, Magnifica Humanitas is the most serious document the debate has produced.