The 24 hours since Pope Leo XIV published his first encyclical, Magnifica Humanitas, have produced something unusual: a papal document being read in earnest by financial and policy capitals as a piece of tech-regulation analysis rather than a piece of theology.
The text addresses governments, parliaments and the executives of the largest AI companies directly, in language that the Holy See has rarely used about a single commercial technology.
The encyclical names artificial intelligence as the present generation’s industrial revolution and argues that without enforceable limits, it will deepen inequality, erode human agency and concentrate power in a small group of firms.
Pope Leo calls for “robust legal frameworks, independent oversight, informed users, and a political system that does not abdicate its responsibility.”
He calls explicitly on states to “disarm AI,” meaning, in the encyclical’s formulation, to remove the technology from purely military and economic interests and place it inside frameworks designed to protect the common good.
The Vatican’s framing of AI as a power-concentration risk is the part being read most attentively in policy capitals. Leo writes that the risk is not only that someone interacting with an AI agent might believe they are talking to a human, but that they might over time lose the desire to seek other people at all.
The encyclical singles out the impact of synthetic content on children and on democratic discourse, and identifies the small number of firms now setting global AI norms as a structural concern in its own right.
The document’s release on Monday was unusual for a papal text in featuring a public conversation with Anthropic co-founder Christopher Olah, who used the Vatican stage to argue that AI cannot be steered by AI labs alone.
That choice of interlocutor was itself read as deliberate: Anthropic has spent the past two years arguing publicly for external oversight of frontier models, and Pope Leo’s decision to give it a Vatican platform put the encyclical inside a particular wing of the existing AI-policy debate rather than opposite all of it.
The encyclical extends a position Leo has been developing since his election. He used his first papal visit to La Sapienza University in Rome earlier this year to denounce AI-enabled warfare and European rearmament, in unusually direct language. Magnifica Humanitas formalises that line.
It also formally supersedes the earlier Vatican “Rome Call for AI Ethics”, which had been criticised, including in our own coverage, for lacking enforcement teeth.
What the encyclical actually shifts is less obvious. Papal documents do not carry legal force outside the Catholic Church, and the AI policy debate in Brussels, Washington and Beijing was already loud before this one.
But Leo’s document gives a moral vocabulary to legislators and regulators who have been groping for one, and the Vatican’s diplomatic reach into governments that the EU AI Act does not touch, parts of Latin America, sub-Saharan Africa and South Asia, is not trivial.
The European Commission welcomed the encyclical on Monday evening; OpenAI, Google and Microsoft offered formal expressions of respect; Anthropic, through Olah, was already on the record.
The next public test is if the Vatican translates the encyclical into specific positions during the UN’s General Assembly AI discussions in September. The Pope has signalled he intends to.
