The pope has used his first encyclical to issue an unusually blunt warning on artificial intelligence, saying AI must be “disarmed” and stripped of the logics that turn it into a tool of domination, manipulation, and automated killing.
Summary
- In his encyclical Magnifica Humanitas, Pope Leo XIV calls for AI to be “disarmed,” comparing its potential to nuclear energy and demanding it be harnessed only for the common good.
- He warns that AI is already fueling conflict, misinformation, and “new forms of slavery,” and insists that delegating lethal decisions to machines is “not permissible.”
- The pope urges world leaders to slow development, build strong legal frameworks, and ensure AI serves human dignity, echoing and sharpening earlier Vatican concerns about autonomous weapons and deepfakes.
In a landmark document released on May 25, Pope Leo XIV presented Magnifica Humanitas, the first major teaching text of his papacy and a sweeping manifesto on AI, technology, and human dignity. He writes that artificial intelligence must now be “disarmed, liberated from logics that transform it into a tool for domination, exclusion, and destruction,” and pointedly compares AI to nuclear power—something that can serve everyone but, left unchecked, can also annihilate entire societies.
The encyclical warns that a global “competition for increasingly powerful algorithms and expansive datasets,” driven by geopolitical rivalry and commercial greed, is pushing humanity toward systems that optimize for control rather than care. In that race, he argues, AI has already begun to deepen global conflicts by accelerating misinformation, amplifying polarisation and lowering the psychological threshold for war when lethal decisions are delegated to code.
“No machine should ever choose to take the life of a human”
Pope Leo’s harshest language is reserved for AI in warfare and state power. Echoing earlier Vatican documents under Pope Francis, he condemns Lethal Autonomous Weapon Systems that can “identify and strike targets without direct human intervention,” calling them a “cause for grave ethical concern” and insisting it is “not permissible” to entrust irreversible, lethal decisions to machines.
Previous Vatican teaching on AI, including the note Antiqua et Nova and Francis’s World Day of Peace message on “Artificial Intelligence and Peace,” already framed autonomous weapons as an “existential risk” that could threaten the survival of entire regions or even humanity itself. Leo explicitly builds on that line, stating that “no machine should ever choose to take the life of a human being,” and calling for international law to ban systems that act as de facto automated executioners on the battlefield or in policing.
The encyclical also targets what he describes as AI-driven “new forms of slavery,” from opaque algorithmic management of workers to exploitative surveillance and deepfake pornography that strip people of control over their image and identity. These dynamics, he warns, risk creating a tiered society in which those who design and own AI systems exert unprecedented power over those who are merely measured, scored, or simulated by them.
A call to slow down—and to regulate hard
Leo’s message is not a blanket rejection of technology, but a demand that AI development be subordinated to human dignity and democratic control rather than the other way around. The encyclical calls for “strong legal frameworks, independent oversight, informed users, and a political environment that does not relinquish its obligations,” warning that governments cannot simply outsource responsibility to engineers or platform CEOs.
He urges world leaders to “slow down” the AI arms race, particularly in domains such as military systems, mass surveillance, and political manipulation, where the incentives to deploy harmful capabilities are strongest. At the same time, he acknowledges AI’s “immense potential” in medicine and social welfare so long as it remains a tool that complements human judgment instead of replacing doctor‑patient relationships or eroding face‑to‑face solidarity in moments of illness and vulnerability.
The encyclical closes by inviting other religions, civil society, and technologists themselves to treat AI not as an inevitable destiny but as a contested field of choices about what kind of civilization humanity wants. In Leo’s framing, the question is not whether AI will transform the world, that is already happening, but whether societies are willing to “disarm” it before it locks them into systems of domination that no one ever consciously chose.
Be the first to comment