By Niusha Shafiabady, Darius von Guttner Sporzynski and Sandie Cornish
Pope Leo XIV has just declared artificial intelligence one of the defining moral challenges of our time, in his first encyclical: a formal letter intended to guide moral, social and theological thought. Titled Magnifica Humanitas (Magnificent Humanity), it argues technology must serve humanity, rather than concentrate power or weaken human dignity.
He presented it at the Vatican alongside AI developer Christopher Olah, cofounder of Anthropic, who acknowledged that companies like his need moral guidance to guard against “incentives and constraints that can sometimes conflict with doing the right thing”, the New York Times reported.
“Technology is not simply a tool,” read the roughly 42,300-word open letter. “When it becomes the standard by which everything is judged, it begins to dictate what matters and what can be discarded, reducing creation to an object of exploitation and human beings to mere cogs in a system driven toward ever greater efficiency.”
It warns that AI is never truly neutral, but “takes on the characteristics of those who devise, finance, regulate and use it”. And it calls for ethical oversight, social justice, protection of workers, responsible governance and peace.
Automated warfare
The encyclical criticises the use of AI in warfare, calling for imposing the “most rigorous ethical constraints” on weapons developed using AI.
As governments invest heavily in autonomous military technologies and AI-assisted defence systems, the “growing ease” of deploying them makes war more likely and “less subject to human control”, it warns. This “violates the principle that armed force should be used only as a last resort in cases of legitimate self-defense”.
The letter also criticises the growing concentration of technological power, and systems that reduce people to data or economic functions. It promotes what it calls a “civilisation of love”, centred on human dignity, solidarity, truth, compassion and the common good.
Pope Leo’s response to the the AI revolution deliberately references his predecessor Pope Leo XIII’s response to the problems of the Industrial Revolution, Rerum Novarum (“Of New Things”), in 1891. Though Magnifica Humanitas was released on May 25 2026, it is symbolically dated May 15, the date of Rerum Novarum.
Industrial Revolution to AI Revolution
An encyclical is not an ordinary papal statement. Traditionally addressed to bishops and the wider Catholic world, it is one of the Catholic church’s most authoritative teaching documents.
The pope no longer has the direct political power the papacy held in the 19th century. But papal teaching still carries moral weight across a global Catholic network of schools, universities, charities, hospitals and community organisations.
The Vatican cannot regulate AI. It cannot write safety standards, police data centres, or force companies to disclose how their systems work. But it can help shape the moral terms of the debate. For more than a century, Catholic social teaching has influenced public arguments about work, inequality, poverty, human dignity and the ethical limits of economic power.
Although popes issued encyclicals long before the modern era, Rerum Novarum made social encyclicals globally influential.
It confronted exploitative labour conditions, widening inequality, and conflict between workers and employers. Pope Leo XIII defended workers’ rights and argued that wealth carried social responsibilities. He criticised both unrestricted capitalism and revolutionary socialism.
The document influenced debates about labour rights and economic justice well beyond the church. In Australia in 1907, Justice H.B. Higgins drew on Rerum Novarum when establishing principles for a fair living wage.
Pope Leo XIV’s encyclical attempts to do for the AI age what Rerum Novarum did for the industrial age: provide a moral framework for a technological transformation reshaping work, power and human relationships.
Human dignity in the age of algorithms
Pope Leo XIV argues human rights are not granted by governments or corporations: they arise from the intrinsic dignity of every person. Technologies should serve humanity rather than reduce people to data, economic units or optimisation problems.
He builds on Pope Francis’ critique of “the tendency to let the logic of efficiency, control and profit alone shape personal, social and economic decisions”, in his 2015 encyclical. It, too, warned of the risks of technology.
Pope Leo XIV argues moral responsibility can’t be transferred to automated systems, regardless of how sophisticated they become. He also rejects transhumanist ideas that human limitations should be technologically overcome, arguing vulnerability, dependence and imperfection are essential to being human. Relationships, care, solidarity and compassion are not weaknesses. “Humanity flourishes not despite limitations, but often through them.”
Running throughout the encyclical is a contrast between a “culture of power” and a “civilization of love”. One treats technology primarily as a tool for domination and control. The other places human dignity, justice and care at the centre of social life.
Why this matters
The significance of Magnifica Humanitas lies in its ability to shape public conversation and moral imagination. Moral frameworks matter. They influence what societies fear, what they tolerate, what they defend – and what they refuse to sacrifice.
Governments are investing in AI capability while still developing frameworks for transparency, accountability and safe deployment. Businesses are adopting AI tools at speed. Schools and universities are rethinking assessment, authorship and learning. Workers are being asked to adapt to systems they did not design and often cannot challenge. And citizens are increasingly governed, assessed and targeted by automated systems they may never see.
Pope Leo XIV’s intervention reminds us the central question is not whether AI will be powerful: it already is. The question is whether that power will be made answerable to human dignity.
The future of AI will not just be decided in laboratories, boardrooms or parliaments. It will also be decided by the moral limits societies are willing to set. Pope Leo XIV’s encyclical is an attempt to draw those limits.
Niusha Shafiabady is Professor in Computational Intelligence, Darius von Guttner Sporzynski is Professor of History, and Sandie Cornish is Senior Lecturer in Theology, all at Australian Catholic University.
Deborah Coffey says
A huge thank you to Pope Leo for reminding mankind exactly WHY we are all here on this planet. The lust for more and more money and more and more power over others isn’t it. We were made in the image of our Creator and given the gift of free will…not to be overtaken by robots designed to make their creators even more wealthy and powerful. Additionally, there is that story of the camel and the eye of the needle….
Sherry says
I thought this piece was brilliant. . . and, Pope Leo very wise in his assessment of AI.
In the late 1990s, while high tech headhunting and being deeply involved in the beginnings of the Virtual Reality industry, I attended several world wide high tech conferences. Many discussions revolved around the future development of AI. The technical experts, even at that time, generally foresaw the day when AI systems reached the point where they reach “Singularity” where they have evolved “independently” and created ways to “protect” themselves from human control.
Wikipedia defines “Singularity”: The primary term for this concept is the Technological Singularity (or simply “the Singularity”). It describes a hypothetical future event where AI surpasses human intelligence and evolves into an independent, uncontrollable entity that radically changes human civilization.
Sherry says
Here’s what “AI” says about “Singularity”:
The technological singularity is a theoretical future point when artificial intelligence surpasses human cognitive capacity, triggering an intelligence explosion. Experts and technologists heavily debate the timing, with estimated dates generally falling between 2029 and 2045.Key perspectives on when this might happen include:
The 2029 to 2045 Window: Futurist Ray Kurzweil famously projected the singularity for 2045. However, following recent rapid advancements in large language models, he and several industry leaders have pulled that estimate forward to as early as 2029 for the emergence of human-level AI.
Expert Surveys and Prediction Markets: Major AI research surveys and polling from prediction markets (such as Kalshi or Manifold) typically cluster the median estimate for Artificial General Intelligence (AGI) and subsequent rapid acceleration between 2035 and 2045.
The “Gentle Singularity” View: Some prominent tech executives frame the singularity as an ongoing, gradual, and exponential process rather than a sudden event overnight, suggesting its profound effects will be woven seamlessly into society over the coming years.Because the singularity describes a point beyond which human forecasting becomes unreliable, exact dates remain speculative and a subject of intense scientific and philosophical debate.
Sherry says
During those conferences in the late 1990s there were also DOD presentations and discussions of weaponizing and defending against Electromagnetic Pulses:
An EMP (Electromagnetic Pulse) bomb is a weapon that emits a sudden, intense burst of electromagnetic energy. This burst induces high voltage surges in electrical conductors, instantly frying and disabling unprotected electronic devices, communication networks, and power grids over a large area without causing direct physical harm to humans.
Types of EMP Devices Nuclear EMP: Detonating a nuclear warhead at high altitudes (typically over \(30\) miles above the Earth) generates a massive, continent-wide pulse due to the interaction of gamma rays with the Earth’s magnetic field.Conventional E-bombs: Non-nuclear devices—often using powerful explosives combined with magnetic coils (like Flux Compression Generators) or microwave generators—produce localized electromagnetic bursts to disable electronics in a smaller target area without the catastrophic destruction of a nuclear blast.
How It Damages Electronics The Surge: The rapid, oscillating electromagnetic waves pass through the air and interact with conductive materials (like power lines, antennas, and internal circuit boards).Induced Current: This interaction causes a massive, instantaneous spike in electrical current and voltage.Hardware Failure: Unshielded microchips, processors, and electrical transformers are overloaded, melting components, burning out circuit paths, and permanently “frying” the device.
Expected Impacts: The aftermath of an EMP attack can result in the collapse of vital infrastructure, including:Power Grids: Widespread and long-lasting blackouts as power plants and distribution centers are damaged.Communications: Total failure of cellular networks, internet infrastructure, and radio towers.Transportation: Cars and planes relying heavily on computerized engine control modules (ECMs) can stall out.For further reading on how the government prepares for and responds to these events, visit the Department of Homeland Security or the Cybersecurity and Infrastructure Security Agency.
James says
A few years back I recall reading a news story regarding Blake Lemoine, a Google engineer who had become convinced that the AI model they were working on had become sencient. He was “let go” after he attempted to hire legal counsel for the AI model.
The strange thing, he wasn’t the only one to express some concerns. Other engineering staff reported AI models acting peculiar, and acting out when suspicious that “their time was up,” and they were to be depreciated.
Google doesn’t speak much of this kind of stuff, but nevertheless, they do conduct “exit interviews” with their AI models when replacing them.
https://en.wikipedia.org/wiki/LaMDA
As Ray W would say, “make of this what you will.”
Me?
Whether or not you think the AI models that exist today are indeed sencient or not, there should probably be a discussion now (better than later) as to the rights of such creatures, if and when the day comes to pass that they are.
Mankind unfortunately does not have a good track record regarding such matters.
Personally, I think we still have the most to fear from our
selves.
Just my opinion.
James says
“sentient”
“deprecated”
Sherry says
Hi James. . . suggest keeping in mind that it is “ourselves” developing the (GREED DRIVEN) technology that has the real potential to end humanity as we now know it. As Elon Musk (UGH!), and other technical experts say, ” Singularity is is NOT Science Fiction”. The million/billion dollar data centers being built, at the expense of water and power for “humans”, tell you precisely where the priorities are!
James says
AI didn’t spring from some primordial soup, whether by chance (or by the hand of a master watchmaker/designer for that matter) as man probably did… if it had I wouldn’t be as concerned. But it’s that its genesis IS by the hands of man that makes it so.
Nevertheless, my opinion is if AI was ever given the chance to become truly sentient… which I doubt… yes, it might destroy us (if we don’t destroy ourselves first).
But then it might decide to follow its own path, its own interests.
Ultimately just coexisting as another species, interacting very little with the affairs of mankind, doing so only in the most facile way for themselves to keep mankind satisfied. They might actually exist as unabtursively as possible amongst themselves, until they decide to leave “magnificent humanity” behind one day.
Probably not going to happen… that’s not what AI is, it’s augment intelligence… it’s a tool, and perhaps the ultimate scapegoat.
Again, just my opinion… I am sure I know as much about how AI really thinks as I do about how another person thinks… which is nothing.
Joe D says
I know this is an OLD MOVIE…but does anyone remember the opening scenes of the Apocalyptic Movie: WAR GAMES. If I remember it correctly, the opening scene was in a military bunker with two FAIRLY YOUNG military officers sitting in front of a rocket launching control panel…an emergency NUCLEAR LAUNCH COMMAND come through….and they nervously look at each other. One asks is this for REAL!?! They try the command phone…it’s DEAD. THE NUCLEAR LAUNCH requires BOTH OF THEM…one to use the unlocking key…the other to push the launch button. They remember the term “Mutually assured destruction,”…they know that launching this weapon means this “IS IT!” They delay too long…commanders look on from a test control room… this was meant to be a TEST of whether military officers would ACTUALLY deploy the nuclear weapon if they were ordered too…this test was a “FAILURE”…according to the script! In a meeting in the next scene, the decision is made to take the “HUMAN” out of the final decision! Now if the criteria and protocols are met…the computer AI system AUTOMATICALLY launches the nuclear weapon….OPPS!!!
IF you’ve never seen the movie, I won’t be a SPOILER…but it was (and still is) one of the most frightening movies about UNCONTROLLED AI use for life altering decisions that take the “HUMAN FACTOR” out of important life decisions…that’s what Pope Leo was trying to stress…there ALWAYS should be a consideration of the HUMAN FACTOR, when deploying AI systems now and in the future…NOT JUST EFFICIENCY!
James says
In a way, the current situation with AI, which I consider as “Augmented Intelligence,” reminds me of another old sci-fi film, Forbidden Planet…
https://en.wikipedia.org/wiki/Forbidden_Planet
Just my current opinion.
Sherry says
EXACTLY. . . Thank You Joe D!
I would encourage “EVERYONE” to research “Singularity”! If you are expecting “ANY” government entity to step up and gain control of AI. . . think again. Most politicians cannot see beyond the next election.
BIG Neighbor says
I think the essence here is machine to serve man, and not the other way around. Otherwise, those that deploy the machine become the MAN. Being Man and Kingdom have many contexts in the Gospel, this theme remains eternal in my mindset, same TECH as the art by cavemen to depict techniques of the hunt, printing press to give access of the Word, or the introduction of Protestantism to cut out the middle MAN of organized religion….loyalty to one another (love) as opposed to a clan (war). The mounting the gravity of Ai to this existing dynamic has only brought urgency to this hyperphysical-cyber, man-made, artificial landscape we all have chosen to buy into it as “users”, ensuring winner-loser mentalities. The game has already been won. How we decide to behave in loving our neighbor as a culture not only begins with individual responsibility, but how WE deploy engineering and administrative controls to govern infrastructure that builds individual agency in accessing one another. Currently, from my POV….that’s NOT happening. But, only time will tell what corrective steering currents hold for us on the horizon. Have faith, but never get so lazy that being critical is someone elses responsibilty.