Magnifica Humanitas ("On Safeguarding the Human Person in the Time of Artificial Intelligence") is the first encyclical of Pope Leo XIV and the first papal encyclical centered on artificial intelligence. It was signed on 15 May 2026 — the 135th anniversary of Leo XIII's Rerum Novarum (15 May 1891) — and released on 25 May 2026. The English text runs about 42,300 words across 51 pages in the official PDF. An encyclical is the most formal mode of Catholic doctrinal teaching, addressed to the universal Church; the Leonine title echoes Rerum Novarum and situates AI within Catholic social doctrine, framing it as a new, cognitive industrial revolution.
The canonical text is published at vatican.va.
Structure
The encyclical comprises five chapters plus an introduction and conclusion. The AI-specific argument is concentrated in Chapter Three ("Technology and Dominance: The Grandeur of Humanity in Light of the Promises of AI"), with the war and autonomous-weapons treatment in Chapter Five ("The Culture of Power and the Civilization of Love"):
- A dynamic approach faithful to the Gospel; the development of social doctrine from Leo XIII to the present.
- Foundations and principles of social doctrine (human dignity; common good; universal destination of goods; subsidiarity; solidarity; social justice).
- Technology and dominance — the technocratic paradigm; what AI is and is not; responsibility, transparency, and governance; transhumanism and posthumanism.
- Safeguarding humanity in transition — truth, work, freedom.
- The culture of power and the civilization of love — including weapons and AI and the crisis of multilateralism.
Summary of argument
The encyclical argues that AI is "cultivated," not "built." "Current AI systems are more 'cultivated' than 'built,' for developers do not directly design every detail, but instead create a framework within which the intelligence 'grows'" (§98), so that "fundamental scientific aspects — such as the internal representations and computational processes — remain, at present, unknown." This "scaffold/grow" language reportedly reflects Anthropic's influence (see Pope Leo XIV).
It holds that AI cannot bear moral agency. AI systems "do not undergo experiences, do not possess a body, do not feel joy or pain… Nor do they have a moral conscience" (§99). They "imitate language, behavior and analytical skills, or even simulate empathy and understanding, but they do not understand what they produce." The "learning" of a model is "a form of statistical adaptation based on data and feedback… [that] does not imply inner growth."
The text rejects the view that AI is morally neutral: "We cannot consider AI to be morally neutral… every technical tool embodies choices and priorities through what it measures, ignores and optimizes, and how it classifies people" (§104). Ethical discernment "cannot be limited to asking whether we are using a system for good or bad purposes; it must also examine how that system is designed and what vision of the human person… is embedded in the data and models."
Power concentration is named as the central danger. "AI tends to amplify the power of those who already possess economic resources, expertise and access to data… small but highly influential groups can shape information and consumption patterns, influence democratic processes and steer economic dynamics to their own advantage" (§108). Data "cannot be left solely in private hands but must be appropriately regulated," and should be treated "as a common or shared good."
On accountability and the pace of adoption, the encyclical states that calling for "prudence, rigorous evaluation and even, at times, a slower pace in adopting AI does not mean opposing progress" (§106). "It is not enough to invoke ethics in the abstract; robust legal frameworks, independent oversight, informed users and a political system that does not abdicate its responsibility are required." It adds that "a more moral AI is not enough if that morality is determined by a few" (§107) — alignment alone is insufficient without the ability to "openly discuss the ethical frameworks involved and subject them to shared standards of social justice."
The document reframes AI through the principles of social doctrine (§109): the common good "means exposing… the new monopolies of AI"; the universal destination of goods implies universal access to technology and education; subsidiarity requires communities able to make choices, not just oversee after the fact; solidarity entails recognizing "the hidden, often exploited workers, who sustain algorithmic systems"; and justice means questioning "who in fact can train these models and who is merely subjected to them."
It calls to "disarm" AI: "Disarming AI means freeing it from the mentality of 'armed' competition… a race for ever more powerful algorithms and larger datasets, driven by the desire to secure geopolitical or commercial dominance" (§110). To disarm "does not mean rejecting technology, but preventing it from dominating humanity… merely regulating it is insufficient; it must be disarmed, welcoming and accessible." Chapter Three closes with a critique of transhumanism and posthumanism (§§115–116) as the "ideological background" that treats the human person "as an object to be optimized" rather than a person called to relationship.
Weapons and AI (Chapter Five, §§197–200)
Chapter Five contains the encyclical's most direct intervention into the Autonomous Weapons debate. It argues that autonomous weapons make war "more 'feasible' and less subject to human control," violating the principle that armed force be a last resort (§197). It states that "it is not permissible to entrust lethal or otherwise irreversible decisions to artificial systems… No algorithm can make war morally acceptable" (§198), because moral judgment "cannot be reduced to calculation."
The text sets out three discernment criteria (§199): identifiable and verifiable personal responsibility; a moral timeframe in which speed never drives irreversible war decisions; and identification and protection of civilians ("any technology that facilitates attacks without seeing the face of human beings lowers the moral threshold of conflict"). It frames three non-negotiable requirements (§200): full traceability and reconstruction of decision-making ("so that accountability… [is] not collapsed into 'the machine'"); lethal-force decisions kept "under effective, self-aware and responsible human control"; and an international shared framework to curb the arms race. Coverage frames this as declaring just-war theory "outdated" for machine-mediated lethal force.
Reception
The encyclical's pro-oversight, anti-concentration posture placed it in immediate public tension with the Trump administration's deregulatory line, drawing pushback from David Sacks on 26 May, the same week as the signed Trump frontier-AI EO. The Vatican is among the few non-governmental, non-industry institutions of global reach engaging frontier-AI governance at this level. Scholars including Green and Herzfeld (see Pope Leo XIV) read the document as institution-formation for the Church and a "moral imperative" offered to industry rather than a rule, with an operative audience of Catholic clergy and the faithful as much as Silicon Valley.
Provenance
The page draws on the encyclical text itself for the quotations above, having previously been built from secondary reporting in Pope Leo XIV. The canonical primary text is the Vatican PDF and HTML edition cited in the lead.
Relationships
- supports: Pope Leo XIV — canonical primary text for the Pope's AI thread; AI Power Concentration — the encyclical's "predominantly private" power-concentration thread; Autonomous Weapons — categorical bar on delegating lethal decisions.
- instance-of: AI Ethics Frameworks
- related: Religious and Civil-Society Voices on AI · AI Labor Disruption · AI Welfare / Model Welfare / Moral Patienthood · Anthropic · Chris Olah · Algorithmic Accountability and Bias Audits
- contradicts: the "technology is neutral; only use matters" framing common in industry advocacy; mandatory-deference-to-developer-autonomy framings of alignment.