Showing posts with label Artificial Intelligence. Show all posts
Showing posts with label Artificial Intelligence. Show all posts

Sunday, September 20, 2026

When a Face Can Lie: What Deepfakes Teach Us About the Self





When a Face Can Lie: What Deepfakes Teach Us About the Self

Savio Saldanha SJ

20 September 2026

DOI - 10.5281/zenodo.22858238



                My reflection on this topic did not begin with a news article. It began several months ago reading about what AI-generated images and deepfakes are doing to people, quietly, in numbers nobody is really counting. Most of what happens this way never becomes a story. This is especially a serious threat in places where the law hasn't caught up to the technology, or was never going to move fast enough to matter to the person it happened to. For every case that gets reported, there are certainly many more that don't. Both men and women are targeted this way, though what's done to women more often carries a specific sexualised cruelty.

                Out of the several cases that I read, one particular case did stay with me while I was working through this, and I want to use it carefully, because a reflection needs somewhere concrete to stand. Georgia Horton (https://bbc.com/news/articles/c75gvyk6d4yo) found out that Ethan Orchard, a man she had known since college and someone she considered a friend, had taken her photographs from social media and used artificial intelligence to turn them into sexually explicit images. He had done the same to two other women he knew. He was eventually convicted and sent to prison. Georgia decided to speak publicly about it afterward, not for sympathy, but because she wanted people to understand that this technology has a dangerous side most people are still treating as a joke. I'm not going to describe more of what was done to her mainly because it doesn't need repeating to make the point. What I want to sit with is something Georgia said plainly: none of this happened because a stranger attacked her in the street. It happened because someone who knew her, who she trusted, decided her face and body could be taken and remade into something she never was and never consented to. The same thing has happened differently to people this same week some of whose stories will never reach a court or a headline.

                That's worth pausing on before any philosophy enters the picture. Whatever follows here isn't an answer to what was done to her, or to any of the others. It's not meant to make any of it smaller, tidier, or easier to file away. But sitting with it long enough, an honest question does form: if a person's face and body can be taken and reassembled like this, entirely without their will, what was actually stolen? I think we can all agree on one thing as of now, that something real was taken. The question is what exactly, and what wasn't.

The self isn't what shows up on screen

                Centuries before any of this was technically possible, Descartes tried a version of the same question in his own mind. Suppose some deceiver could fabricate every single thing I see and hear. Could I trust anything? He found one thing that survived: even a total deception needs someone present to be deceived. I think, therefore I am. Somebody has to be home for a lie to land on. Here is what Georgia's case actually clarifies about that old argument. Descartes was only ever certain about his own case, from the inside. The cogito was never a test for detecting someone else's presence behind a likeness — it says nothing about whether an image of a person carries that person's thought, consent, or will. What was done to Georgia is proof of exactly that gap: a convincing image can be built entirely without the will of the person it depicts, in direct violation of that will. They used her likeness, but nothing in them was her choice, her thought, her consent. That distinction changes nothing about what she suffered. You might think there's nothing more to say here: a person's likeness was taken, full stop. But it's worth being precise about what was and wasn't touched. It does name something true and worth holding onto: whatever was in those images, it was not her lived presence, her will, or her consent.

What actually can't be copied

                Buddhist thought does not affirm a fixed, permanent, independent self behind our experience — but that isn't a denial that a person's experience, or the harm done to it, is real. It still insists that something real is happening in a person: a stream of actual, lived experience, one moment shaping the next, carrying memory and consequence forward. That continuous, lived stream is not something an image can hold, no matter how well it mimics a face. The pictures Orchard made had no experience, memory, will, or consent of their own. They were surfaces built from her data, not living centres of experience — with none of her lived experience, will, or consent actually inside them.

So in that narrow but real sense, what was copied was never Georgia. What was violated instead was something else entirely, and it deserves to be named honestly rather than resolved away by a philosophical technicality.

The wound is real even where the self isn't touched

                Here I want to be direct, because a comforting-sounding argument can do real damage if it's used carelessly. It would be insensitive to tell Georgia, ‘don't worry, your true self was never in those pictures,’ as if that settles anything. It doesn't undo what strangers said about her online. It doesn't undo the violation of being sexualized without consent, by someone she trusted, for other men's entertainment. Those are real and specific harms done to a real person and no metaphysical distinction erases them.

                What I think is actually true is that both things hold at once, without one cancelling the other: nothing essential about who Georgia is was ever in Orchard's power to touch or remake, and at the same time he still did something to her that a court rightly called criminal, that caused her real and lasting harm, and that deserves anger, not just analysis. A philosophy that only offers the first half of that sentence isn't consoling anyone rather it's avoiding her.

A face that makes a claim on you

                The philosopher Emmanuel Levinas spent much of his life on one idea worth slowing down for. When you actually encounter another person's face, not glance at an image of it, but meet it, something happens before you've judged anything about it. The face itself makes a claim on you, Levinas said: something like do not reduce me to what you can see, I am not yours to use. That claim doesn't come from any feature you could photograph. It comes from the fact that a real person always exceeds any image taken of them.

                This is exactly what Orchard's images faked, and exactly what they could never actually deliver. Every man who looked at those pictures and treated them as an invitation was responding to something that made no genuine claim on them as an encounter with Georgia, because nothing was actually there to be encountered — only data arranged to look like an encounter. That is where I think the deepest wrong in this crime sits, underneath even the images themselves: it manufactured the feeling of having met a woman, judged her, and been given license by her, a whole moral event, when in truth no meeting had happened, and no license was ever hers to give. Every one of those persons owed Georgia the restraint that a real face demands, and everyone who didn't give it failed a test the fake image was specifically built to help them fail.

Being known, not just being seen

                Catholic theology has never located a person's worth in physical resemblance. Genesis says human beings are made ‘in the image and likeness of God’ (Gen 1:26–27), and the tradition has always read that as naming a capacity for relationship, intellect, and freedom, not a physical template that could be copied or corrupted by copying. The Psalmist says it more intimately still: ‘You knit me together in my mother's womb... my frame was not hidden from you’ (Ps 139:13,15). Before Orchard, before any camera, before anyone else's gaze fell on her at all, Georgia was already known this way completely by someone no lie could deceive. Jeremiah hears it as a word spoken before he even existed: ‘Before I formed you in the womb I knew you’ (Jer 1:5). Isaiah makes the same claim in a different register, and it is the verse this whole reflection keeps returning to: ‘I have called you by name, you are mine’ (Is 43:1). That is not a description of how anyone looks. It is a description of how a person is held.

                Paul pushes the same conviction into the body itself, which matters directly here, since it was her body that was stolen and misused: ‘Do you not know that your body is a temple of the Holy Spirit... you are not your own’ (1 Cor 6:19–20). The body is not incidental to the person in Catholic thought the way it might be in a purely Cartesian picture. But it is not reducible to an image of it either. The Catechism holds both of these at once: the human person, ‘willed for its own sake,’ bears within it ‘the divine image’ precisely in the soul's spiritual powers of intellect and will (CCC 1700–1705), and yet the Church has always taught that a person's reputation is not a trivial thing to protect: to expose someone unjustly, to falsify what they've said or done, to let others believe something untrue about them, are named directly as sins against truth and against the neighbour (CCC 2477–2478). What was done to Georgia is not new under a new name; the Church has always had words for it. It simply now happens at a scale and a speed her tradition never had to reckon with before.

                Vatican II said something that reads as though it were written with this exact harm in mind, decades before the technology existed to commit it. Gaudium et Spes lists, among the acts that ‘poison human society’ and ‘do more harm to those who practice them than those who suffer from the injury,’ whatever ‘is offensive to human dignity’ (GS 27). And in 2024, the Church finally had to say it in language that names this precisely: the Dicastery for the Doctrine of the Faith's declaration Dignitas Infinita devotes its closing section to what it calls ‘digital violence,’ naming directly the way ‘new forms of violence are spreading via social media,’ including ‘the sharing of images of a sexual nature without consent’ (DI 61–62). This isn't an old doctrine stretched to cover a new problem. It's the Church watching this exact thing happen and finding it needed a new name for a very old wound: dignity attacked not by a fist or a word, but by a fabricated image, at the speed of a share button. Orchard could violate Georgia's privacy, safety, reputation, and trust — the Church has never pretended those losses are small. He could not cancel the dignity she carries as a person made in the image of God.

A fullness that cannot be subtracted

                The Īśā Upaniṣad reaches a strikingly similar insight by an entirely different road, centuries before either Descartes or the Gospel:

ॐ पूर्णमदः पूर्णमिदं पूर्णात्पूर्णमुदच्यते।
पूर्णस्य पूर्णमादाय पूर्णमेवावशिष्यते॥

‘That is full, this is full; fullness comes forth from fullness. Even when fullness is taken from fullness, fullness alone remains.’

                This is obviously not identical to the Christian doctrine of imago Dei, and the two traditions shouldn't be collapsed into one another. Yet both refuse the same reduction: that a person is exhausted by whatever can be seen, displayed, or manipulated. Something enormous was taken from Georgia. This verse doesn't say otherwise. The theft was real, but it never exhausted the person it tried to define. In other words, there is more to a person than what we see.

The one workable thing

                Most writing about deepfakes stops at feeling uneasy about the technology, whoever the victim happens to be. Here is something concrete, and it follows directly from Levinas's point above: before looking at, believing, or sharing an image that claims to show a real person, ask one question. Has this person actually consented to be seen this way, or am I about to treat a stolen likeness as if it were a real encounter with them? If there's any doubt, don't look, don't forward, don't engage. That single refusal is small, but it's exactly the discipline Levinas asks of anyone who meets a real face: to answer to the person, not just to the image.

What's left unresolved

                Georgia's case did end with a conviction, which is more than most victims of this get, and more than most will ever get. But she was clear that the law still isn't enough, even where the law exists and works. The tools to do what was done to her are still freely available, and she said so herself. Somewhere now, the same thing is being done to someone whose name will never surface, in a country where no one will be charged, where there may be no law to charge anyone under at all. I cannot let one filmed, named suffering matter more than the thousands who may have been affected the same without a voice or a name the world will ever know. That would be its own kind of selective outrage, the very thing I have tried to hold myself against throughout this piece. Georgia is not the point, but a door through which the point became impossible to avoid.

                Knowing that no theft could reach the fullness of who any of them are doesn't undo the months of fear, or the fact that people looked at fabricated images and felt entitled to comment on what they'd like to do to another person's body. Theology can say truthfully what wasn't lost. It can't yet say how, what actually was lost, safety, trust, the ordinary freedom of existing as an image online without fear, gets given back. That isn't something an argument resolves. It's rebuilt, slowly, through better law where law can reach, through people refusing to look, and through communities choosing to stand with the person rather than with the image someone else made of them, wherever in the world that person is, whether or not we ever learn their name.

 A Pastoral Afterword

                If these reflections leave you unsettled, uneasy, or even defensive, pause before judging that reaction. Discomfort is not always a sign of error; often it is the first movement of conscience waking up.

                To anyone reading this who has been a victim of this technology or in any form however it reached you: nothing above was written to explain your pain away and nothing here asks you to feel less than you do. You are more than the image that was made of you. Not as a comforting slogan, but as the plain truth this whole reflection has tried to hold onto: whatever was taken, copied, or twisted was never the whole of you, and it was never going to be. That doesn't undo what happened. It also doesn't need to be resolved today, or explained fully, or forgiven on any timeline but your own. If what you need right now is simply to be believed and not analysed, that is enough, and this piece was never meant to ask more of you than that.

                To anyone reading this who recognises themselves differently, someone who has used this technology, looked at what it made, shared it, laughed at it, or turned to it in a private moment they'd rather not name, even here, this is not an accusation meant to shame you into silence. Shame closes a person off from exactly the kind of honest self-examination that could actually change something. What's offered instead is the same as always: an invitation to look again, gently, at something conscience may already be asking you to look at. Ignatian spirituality holds that God rarely speaks in the loud register of accusation. More often it's the quiet stirring that won't quite let a person settle, the unease that returns after the screen is off, the small resistance to doing again what was done before. That stirring is not your enemy. It may be the most trustworthy voice in the room.

                You are not asked here to make a public confession, to relitigate every choice you've made online, or to carry a guilt heavier than what's actually yours. You are asked only to listen, to the person a fabricated image can never fully replace, and to whatever it is in you that already suspected this technology was never as harmless as it felt in the moment.

                Remain with the unease, whichever side of it you find yourself on. Pray with it. Let it teach you. At the crossroads of conscience, God does not coerce; God accompanies. And for both the one who was wronged and the one still learning what they've done, that accompaniment is itself a sign that love, when it finally roots itself in truth, has not failed. It is only beginning to be purified.


Saturday, December 20, 2025

Should a Machine decide to kill: Catholic Conscience and the Ethics of Unmanned Combat Vehicles


Savio Saldanha SJ

DOI: 10.5281/zenodo.17993534

20-12-2025 

            I recently watched a movie, about some Unmanned Aerial Vehicle operators’ dilemma. They are ordered to target a building where a terrorist is hiding. While outside this building a young girl is selling bread. The operators face a dilemma, fire the missile and kill the innocent girl with the terrorist or let the terrorist escape. I could feel their tension within me as I watched and hoped that the girl would be somehow miraculously saved. But after the movie, I began reflecting on the news reports about the UCV’s.

            Unmanned combat vehicles (UCVs) expose a fault line in contemporary conscience: they promise “clean” warfare while intensifying the moral distance between the one who kills and the one who dies. At this crossroads, Catholic moral theology and recent papal teaching converge on a clear point: no machine may be allowed to decide to take a human life, and any system that obscures or fragments human moral responsibility is ethically suspect, even when it operates inside a “just war” framework. The question of my reflection, “is it ethical, using robots to kill human beings?”

From Predator Drones to “Killer Robots”

            UCVs emerged in the late 20th century as tools for reconnaissance and targeted strikes, exemplified by the U.S. Predator and Reaper drones used from the Gulf conflicts to Afghanistan and Iraq, and later in theatres like the Armenia–Azerbaijan and Russo‑Ukrainian wars. What began as remotely piloted aircraft has expanded into an ecosystem of unmanned ground vehicles, kamikaze drones, and sea‑borne systems increasingly coupled with artificial intelligence for detection, tracking, and engagement.

            At each step, two promises drive their deployment: greater protection for one’s own soldiers and “surgical” precision in targeting, ostensibly reducing civilian casualties. At the same time, these systems deepen a psychological and moral distance: operators may be thousands of kilometres away, or, the machine may operate in fully autonomous modes, humans may be removed from real-time decision-making altogether, turning battlefields into laboratories for live testing of algorithms.​

            The fact that unmanned combat systems cost a fraction of the previously conventional systems this makes it possible for small countries with limited defence budget to field a sizeable fleet of UCV’s. Unmanned aerial systems operating in swarms can overwhelm enemy air defences and cause severe damage to their military and other infrastructure facilities. UCV’s can operate in high-risk zones thus reducing own casualties while inflicting serious casualties on the other side. These entire factors make the UCV’s a lucrative deal.

A New Technological Temptation

            Philosophically, the ethical problem is not technology as such but a new configuration of power, distance and uncertainty. Hans Jonas, was a German American philosopher. While reflecting on technological power in general, he argued that the unprecedented scale and irreversibility of our actions demand a new “imperative of responsibility”, which is act so that the effects of your action are compatible with the continued, genuinely human life of humanity on earth. For Jonas, not everything that can be done technologically ought to be done; technological possibility must be constrained by an anticipatory ethics that takes seriously the worst‑case scenarios, a “heuristics of fear” that refuses naïve optimism about neutral progress.​

            The ethical challenge posed by UCVs is not technology itself, but what technology enables when it reshapes responsibility. Hans Jonas’s Imperative of Responsibility offers a crucial philosophical lens here. Jonas argued that modern technological power has outpaced traditional ethical frameworks. Jonas’s thought resonates deeply with Catholic moral theology. His insistence on restraint, foresight, and responsibility parallels the Church’s emphasis on prudence, moral accountability, and the common good.

            When we apply this principle to autonomous or semi‑autonomous weapons, Jonas’s approach highlights two dangers: first, that the chain of responsibility becomes so distributed (programmers, commanders, operators, political authorities) that no one decisively “owns” the lethal decision; and second, that the very success and efficiency of these systems seduce societies into normalising permanent, ‘low visibility’ warfare. What looks like rational, data driven targeting can conceal a profound moral abdication, since algorithms cannot bear responsibility or suffer remorse, yet their opaque decisions shape life and death on the ground.

            At this point, I believe it is important to clarify my stance. Unmanned combat vehicles are not intrinsically immoral; rather, the ethical evaluation depends on the manner of their use and the degree of human moral responsibility retained in their deployment. These Unmanned systems are a boon when they are used for non-lethal activities like monitoring traffic, delivering medical aid and essential supplies to remote places and some are even configured for medical evacuation. My argument is therefore not anti-technology, but ethically critical. As Hans Jonas insists, the problem is not technological capability as such, but moral abdication — when human agents allow responsibility for life-and-death decisions to be obscured or displaced by technical systems.

Just War, Human Agency, and the Limits of Delegation

            Catholic moral tradition does not approach war with naïveté. Classical ‘just war’ theory — articulated by Augustine, Aquinas, and reaffirmed in the Catechism of the Catholic Church (§§2307–2317) — insists that even in a defensive war, combatants are bound by moral principles. Among these, ius in bello principles such as discrimination and proportionality require concrete moral judgment in each act of lethal force.

            Contemporary ethicists analysing autonomous weapons argue that systems which remove or radically weaken human control over individual lethal decisions cannot in practice reliably satisfy these principles, because algorithms cannot adequately interpret context, intention, surrender, or non-combatant status in the way moral agents must.

            A detailed philosophical analysis by Wendell Wallach and Allen Colin concludes that “human‑out‑of‑the‑loop” weapons — those that select and engage targets without real-time human intervention — are “highly morally problematic”. This is precisely because their design and use impede human agents’ ability to exercise morally informed judgment, and thus amount to an abdication of responsibility. Even where there remains a nominal “human in the loop,” the speed, complexity, and opacity of AI‑driven targeting can according to Robert Sparrow reduce the human role to rubber‑stamping, undermining the requirement that each act of lethal force be a genuinely personal, accountable decision.​

            Catholic ethics does not claim that unmanned platforms are intrinsically immoral. Rather, it insists that lethal force is morally non-delegable. When technology undermines the capacity of human agents to exercise responsible judgment, it exceeds the moral limits of delegation.

The Magisterium’s Emerging Witness

            Over the last decade, the Holy See has consistently raised alarms about lethal autonomous weapons systems (LAWS), often popularly called “killer robots,” explicitly including armed drones and unmanned vehicles in this category. Vatican representatives to the United Nations have warned that removing human agency from the moral equation is problematic not only ethically but also for the foundations of international humanitarian law, since “autonomous weapons systems cannot be considered as morally responsible entities.”

            Pope Francis had repeatedly sharpened this concern. In his interventions on artificial intelligence and war, he insisted that no machine should ever be allowed to choose to take a human life. Thus calling for the development and use of lethal autonomous weapons to be reconsidered and ultimately banned. In 2024 he addressed world leaders, stressing the need for “ever greater and proper human control” and warning that AI lacks the human capacity for moral judgment and therefore must not be entrusted with lethal decision-making. This magisterial trajectory is not a marginal footnote to Catholic social teaching; it flows from a consistent defence of human dignity, the primacy of conscience, and the demand that technological progress be subordinated to integral human development and the common good.

Conscience at the Console: The Drone Operator’s Dilemma

            From a pastoral standpoint, Church voices have begun to recognise that unmanned warfare creates a new kind of combatant whose battlefield is a screen. Vatican officials have noted that drone operators and those involved in deploying unmanned systems often lack adequate formation and time for moral discernment, even as their split‑second decisions affect lives far away, with psychological and spiritual consequences for both victims and operators.

            There is a double “anesthesia” at work here. On one side, geographical and sensory distance can dull empathy: there is no blood, only pixels; no cry, only data. On the other, institutional distance fragments responsibility: engineers, commanders, analysts, and politicians can each tell themselves they merely played a minor technical role, while the system as a whole carries out lethal actions without any single conscience fully confronting their gravity. For a Christian, this runs directly counter to the vocation to see and respond to the concrete face of the other, especially the vulnerable enemy who remains, even in war, a bearer of the imago Dei.

Is It Ethical to Use Robots to Kill?

            This brings me back to my initial question - “Is it ethical, using robots to kill human beings?” This cannot be answered in the abstract, as though there were a single switch to flip between “ethical” and “unethical.” Catholic moral theology pushes us to distinguish levels:

  • If “robot” means a system that autonomously selects and kills targets without meaningful, responsible human control, current magisterial teaching and serious philosophical reflection converge toward a negative answer: such systems should not be developed or used, and should be subject to a binding international ban.
  • If “robot” means an unmanned platform (air, land, sea) still under robust human moral agency, then the perennial criteria of just war — just cause, right intention, last resort, proportionality, discrimination — still apply, and the question becomes whether such platforms actually help or hinder compliance with these criteria in practice.​

            Yet even in the second, more nuanced case, there remains a deep unease in the Christian conscience. The more warfare becomes asymmetrical, remote, and technologically mediated, the easier it becomes for powerful states to wage low risk perpetual conflicts with minimal domestic political cost and minimal existential exposure of their own soldiers. Jonas’s ethics of responsibility, Pope Francis’s appeals for a ban on lethal autonomous weapons, and the Vatican’s insistence on non‑delegable human agency all point in the same direction: technological sophistication does not lessen the gravity of killing; it heightens the demand for moral scrutiny and self‑limitation.

A Conscience at the Crossroads

            Writing as a theology student and Jesuit scholastic my own context sharpens this debate. I stand between at least three pressures: a global South that often bears the brunt of “remote” wars and experimental weapons, a Western context in which high‑tech security discourses are taken for granted, and an ecclesial tradition increasingly vocal about the non‑negotiability of human moral agency in the use of force. This is not a purely theoretical knot; it intersects with the lives of families fleeing drone‑shadowed skies, with soldiers and operators wrestling with guilt, with policy debates that risk drifting far ahead of public moral reflection.​

            As a Jesuit, I feel that in this situation Ignatian spirituality offers a way forward. Ignatian discernment can offer a way of resisting both technological inevitability and moral numbness in the age of unmanned war. Instead of treating UCVs as neutral tools whose use is determined only by strategic necessity, discernment asks: what is this technology doing to my desires, my imagination and my capacity to be moved by the suffering of concrete persons? In the Ignatian tradition, one is invited to “compose the place,” to place oneself prayerfully in the concrete scene — here, the drone feed, the targeted house, the girl selling bread — and to notice the interior movements that arise: consolation that draws toward reverence for life and justice, or desolation that manifests as indifference, fascination with power, or rationalisation of avoidable harm. Examined in this light, decisions about designing, authorising, or operating UCVs cannot be reduced to technical risk assessments; they become moments of encounter with the Crucified in history, who identifies himself with the victims of “clean” warfare as well as with soldiers whose consciences are strained to breaking point. Ignatian discernment thus calls Christians involved in these systems — engineers, officers, chaplains, policymakers — to a slow, honest scrutiny of spirits, so that choices about “robotic” killing are made, if at all, under the sign of the poor and crucified Christ rather than under the seduction of efficiency, distance, and fear. Although it might feel cliché or impractical to some, I sincerely believe that it is the only manner in which human conscience and responsibility can act in justifiable manner in the usage of the UCV’s.

            From this crossroads, a Christian response might be framed in three movements. First, a clear “no” to machines deciding who lives and who dies — an ethical red line voiced both by Jonas’s imperative of responsibility and by Pope Francis’s call that no machine should ever choose to take a human life. Second, a rigorous, honest examination of whether current patterns of unmanned warfare are truly serving just peace or merely lowering the threshold for resorting to force, especially against populations with little power to respond. Third, a renewed commitment to form consciences — of engineers, military leaders, policymakers, and even religious leaders — capable of resisting the seduction of “clean” killing and insisting that human lives, even enemy lives, are never reducible to targets in a dataset.​

            At this intersection of drones and doctrine, algorithms and ethics, my reflection becomes a question addressed not only to the whole Church but to the entire humanity: will our conscience allow itself to be automated, or will it reclaim the slow, costly, deeply human work of responsibility in the age of unmanned war?


Bibliography

Aljazeera. “Pope Francis Calls for Ban on ‘Lethal Autonomous Weapons’ at G7.” 14 June 2024.

Catholic Culture. “Pope, at G7 Summit, Calls for Ban on Lethal Autonomous Weapons.”

Catholic News Agency. “Vatican Again Calls for a Moratorium on Killer Robots.”

Catholic Philly. “Autonomous Weapons Systems Threaten Peace, Says Vatican Official.” 27 March 2019.

Holy See Mission in Geneva. “Technology Should Better Human Life, Not Take It.” 17 September 2025.

Jonas, Hans. The Imperative of Responsibility: In Search of an Ethics for the Technological Age. Chicago: University of Chicago Press, 1984.

​National Catholic Register. “Drone Wars: The Morality of Robotic Weapons.” ​

Pope Francis. Address on Artificial Intelligence and Peace, G7 Summit, June 2024 (and subsequent appeal: “Reconsider the Development of Lethal Autonomous Weapons”). Vatican News. 9 July 2024.

Stop Killer Robots Campaign. “Statement to the UN General Assembly First Committee on Lethal Autonomous Weapons Systems.” 13 October 2020.

UN, Holy See Statements. “Emerging Technology at the Service of Humanity: Called to Be Peacemakers.” Catholic Bishops’ Conference of England and Wales, 21 May 2024.

Wallach, Wendell, and others. “Autonomous Weapons and Moral Responsibility.” In The Oxford Handbook of Ethics of War, Oxford University Press, 2016. 

 

Saturday, November 29, 2025

A Crossroad Between Bans and Discernment: Forming Conscience for the Age of AI




Savio Saldanha SJ

https://doi.org/10.5281/zenodo.17742128

29-11-2025


Introduction

            Artificial Intelligence (AI) has shifted from abstract concept to everyday reality, affecting education, research, and even personal and spiritual growth. As educational institutions respond with strict rules and bans on AI’s use, there is growing concern that technology is advancing faster than regulation, creating a sense that we are fighting a losing battle. After reading recent papal documents and philosophical literature pertaining to AI, I think this development calls for a more integrative approach: one that promotes understanding, ethics, and discernment in the use of AI, especially among youth and researchers. As we stand on the crossroad of the conscience we face a dilemma of the ethical and moral use of AI by students and youth and the limits for its usage.


The Problem

            Currently, many institutions aim to restrict or eliminate AI use among students, fearing it will replace genuine learning with automated shortcuts. However, as AI becomes more advanced—capable of generating essays, solving complex problems, and evading detection—these methods are increasingly ineffective. This leads to frustration on all sides, eroding trust, stifling curiosity, and weakening genuine intellectual formation.


The Solution

            The future lies not in fighting AI, but in fostering a mature, ethical, and reflective engagement with it. By drawing on Catholic teaching and philosophical inquiry, educators and leaders can guide students towards a responsible, creative, and truly human integration of AI.


AI as a Tool: Neutral by Nature, Ethical by Use

            Pope Francis emphasized that AI is a tool—one that can advance knowledge, democratize education, and serve humanity (Pope Francis, 2024). Like all tools, it carries no moral quality by itself. Its effects depend entirely on how people use it: for good or ill, to foster justice or inequality, to create genuine understanding or simply shortcut effort. Pope Leo XIV likewise notes the double-edged nature of AI, affirming that while “AI is above all else a tool,” its ethical value lies in intention and use (Pope Leo XIV, 2025).


Centrality of the Human Person

            The Church’s teaching places human dignity at the center of all technology. Pope Leo XIV writes that AI must be assessed “in light of the integral development of the human person and society,” not just on material outcomes (Pope Leo XIV, 2025). Pope Francis cautions that even sophisticated technology should not undermine the human capacity for moral decision, reflection, and authentic encounter (Pope Francis, 2024). Both teach that any use of AI must support—not replace—uniquely human creativity, judgment, and critical thought.


Insights from the Philosophy of AI

            The philosophy of AI, as outlined by Müller (2024), not only helps clarify what AI is and is not, but also shows how reflection on AI and reflection on the human person belong together. Müller begins from the classical research program launched at Dartmouth in 1956, which conjectured that every aspect of learning and intelligence could in principle be precisely described and simulated on a machine. In this sense, “Classical AI” is a research project aimed at building computer-based agents that genuinely have intelligence, and it stands in continuity with the well-known distinction between “strong AI” and “weak AI.” Strong AI maintains that an appropriately programmed computer literally has a mind and cognitive states, while weak AI treats computer systems as powerful tools for simulating mental processes and for studying the mind without claiming that the machine itself understands. This distinction resonates with Searle’s “Chinese Room Argument,” which suggests that rule-based symbol manipulation, even when behaviorally successful, is not yet genuine understanding; from a theological standpoint, this confirms that intentionality and consciousness—and therefore moral responsibility—remain rooted in the human person rather than in the artifact.

            Müller then contrasts this classical, ambitious understanding of AI with what he calls “Technical AI”: a family of concrete methods in computer science—search, probabilistic reasoning, expert systems, control engineering, machine learning, and so on—used to build systems for perception, modelling, planning, and action. Here AI is not a claim about minds, but a toolbox for constructing systems that behave intelligently in restricted domains. Since around 2015, the rise of deep machine learning, fuelled by massive data and computing power, has dramatically increased the performance of such systems in translation, text generation, games, vision, and autonomous driving, sometimes surpassing human capabilities in specific tasks. Yet Müller stresses that this success does not settle the philosophical question of intelligence itself; it only shows that certain forms of intelligent behaviour can be produced by non-human, non-conscious mechanisms. Theologically, this supports a nuanced view: AI can exhibit impressive capacities without thereby becoming a subject of rights, duties, or grace, because its “intelligence” is instrumental rather than existential.

            Because of these two strands—classical and technical—Müller argues that the philosophy of AI must address three Kantian questions: What is AI? What can AI do? What should AI be? He proposes an “AI philosophy” that does not merely apply pre-existing concepts to a new object, but allows the very concepts of intelligence, agency, and normativity to be re-examined in light of AI systems. For example, work on the Turing Test shows how operational criteria for “thinking” can shift public language, even if they do not resolve deeper metaphysical issues about consciousness. At the same time, debates about goals and values in AI highlight a crucial limit: current systems exhibit remarkable instrumental intelligence (they are very good at finding means to given ends), but they lack genuine metacognitive reflection on which goals are worth pursuing and why. Müller notes that without such reflection on the goodness and relevance of ends, AI cannot be a full moral agent, and talk of “machine ethics” in a strong sense is misleading.​

            This analysis dovetails with Catholic concerns about “algor-ethics.” If AI systems, even highly sophisticated ones, cannot autonomously ground or revise their own goals in the light of truth and the good, then they must remain embedded within human practices of discernment, responsibility, and virtue. Pope Francis’s call for ethical frameworks for AI can thus be deepened by Müller’s claim that normative reflection is not an optional “add-on” but an elementary part of any genuinely rational life-form. In human beings, this reflective capacity is tied to conscience, practical wisdom, and an openness to transcendence; in machines, by contrast, the selection and evaluation of goals must ultimately be designed, monitored, and judged by persons. The philosophy of AI therefore reinforces a central intuition of Catholic moral theology: intelligent artefacts may transform the conditions of action, but they do not displace the primacy of human agents, whose freedom and moral growth remain at the heart of any authentic “ethics of AI.”​


Papal concerns about “outsourcing” formation to AI

            Recent interventions by Pope Leo XIV deepen this educational perspective by explicitly addressing the temptation to let AI “do our homework” in place of real learning. Speaking to students, he acknowledges that AI can be a powerful aid for study but insists it must never replace the hard work of thinking, judging, and creating for oneself, because these are precisely the activities through which persons grow in freedom and responsibility. In this view, AI belongs to the order of tools, whereas wisdom and moral discernment arise only through the engaged, embodied exercise of human intelligence in relationship with others and with God.

            At the same time, the Pope does not reject AI as such; he calls it “one of the defining features of our time” and urges educators and parents to guide young people toward uses of AI that genuinely help and do not hinder their human development. The decisive question is not whether AI is present in schools, but whether its use forms or deforms students: does it cultivate intellectual honesty, patience, and collaborative learning, or does it encourage passivity, plagiarism, and isolation. This resonates with the broader Catholic insistence that technology must always be evaluated in light of the dignity of the person and the integral growth of children and adolescents, who are particularly vulnerable to the allure of effortless solutions.

            From a philosophical and pastoral standpoint, the papal warning against delegating one’s homework to AI can be read symbolically as a warning against outsourcing the very struggle that makes education transformative. If students learn to treat AI as a substitute for their own judgment and creativity, they risk hollowing out the interior capacities—attention, critical reflection, moral imagination—that Catholic tradition associates with the formation of conscience. By contrast, when AI is used transparently and critically, as an instrument that supports research and reflection without replacing them, it can become an ally in precisely the “intergenerational apprenticeship” that the Church envisions.


The Role of Education: From Control to Formation

            A recurring theme in both papal and philosophical sources is the importance of education as formation, not just transmission of skills or facts. Pope Leo XIV calls for an “intergenerational apprenticeship” so that young people can learn to integrate technology wisely into their lives (Pope Leo XIV, 2025). Education should develop students’ responsibility, discernment, and creativity, equipping them to use AI for real growth—intellectually, morally, and spiritually—rather than as a means of shortcutting learning or escaping effort (Pope Francis, 2024; Müller, 2024).

            Pope Francis warns against technophobia and calls for dialogue—across cultures, generations, and disciplines—to ensure AI serves the common good. He advocates a shared ethical foundation (“algor-ethics”) and stresses the need for healthy politics to direct technological change towards justice, inclusion, and the flourishing of all (Pope Francis, 2024).


Personal Reflection and Practical Recommendations

            As we stand at a crossroads of conscience regarding the ethical and moral use of AI by students and youth, the dilemma often appears stark: should AI be banned altogether or tightly limited, and who has the authority to fix and enforce those limits. Witnessing the rigid enforcement of AI bans in many institutions, it seems increasingly likely that such efforts will fail in the long term, not only because the technology will outpace policing, but because a purely prohibitive strategy neglects the deeper formation of judgment that both philosophy and Catholic teaching demand. The categories developed by Müller help here: if much of what is now called “AI” is in fact “technical AI”—powerful but limited methods for perception, modeling, and decision-support—then treating these tools as if they were already quasi-personal agents to be excluded altogether risks confusion and fear rather than clarity.

            A more coherent response is to move from prohibition to formation, from mere rule-enforcement to an “intergenerational apprenticeship” in the wise use of technology, as Pope Leo XIV suggests. If, as Müller argues, current AI systems exhibit at most instrumental intelligence—remarkable skill in finding means to given ends, but no genuine reflection on which ends are good or just—then the responsibility for setting and evaluating goals necessarily remains with human agents. This implies that institutions, families, and Church communities cannot abdicate discernment to algorithms, nor to external regulators alone; they must themselves cultivate the virtues and criteria by which AI use is judged. In this perspective, the key question is not simply “how much AI is allowed,” but “how do we form persons who can use AI without outsourcing the inner work of thinking, choosing, and taking responsibility.”

            Practically, this means encouraging openness, integrity, and critical reflection on AI rather than secrecy and evasion. Policies will still be needed—there must be some boundaries on plagiarism, data misuse, and academic dishonesty—but these norms should be embedded in a broader pedagogical project that teaches students not only what to avoid, but how and why to use AI well. Drawing on Pope Francis’s call for “algor‑ethics,” educators can invite students to ask in each concrete case: does this use of AI support or undermine my own learning, my relationships, and the dignity of others. By engaging youth in such questions, institutions help them move from passive users of opaque systems to discerning subjects who not only understand the technical limits of AI but also its moral implications.​

            In this light, I conclude that, instead of asking whether to ban or permit AI in the abstract, our central task is to shape a culture in which AI is integrated into education in a way that preserves the primacy of human intelligence, conscience, and community. Those who “decide the limits” of AI use—teachers, parents, Church leaders, and students themselves—should be seen not primarily as regulators but as co-responsible participants in a shared work of formation. By teaching young people to collaborate with AI without surrendering their capacity for wonder, critical thought, and moral responsibility, institutions can foster maturity, wisdom, and resilience—qualities that the philosophy of AI identifies as properly human, and that Catholic spirituality recognises as the fruit of grace working through human freedom.


References

  • Müller, V. C. (2024). Philosophy of Artificial Intelligence: A Structured Overview. In: Smuha NA, ed. The Cambridge Handbook of the Law, Ethics and Policy of Artificial Intelligence. Cambridge Law Handbooks. Cambridge University Press; 2025:40-58. DOI: https://doi.org/10.1017/9781009367783.004
  • Pope Francis. (2024, June 14). Address at the G7 Session on Artificial Intelligence, Borgo Egnazia, Puglia. The Holy See.​
  • Pope Leo XIV. (2025, June 19). Message to participants in the Second Annual Conference on Artificial Intelligence, Ethics, and Corporate Governance, Rome. The Holy See.​
  • Pope Leo XIV. (2025, November 21). Address to young people at the National Catholic Youth Conference, Indianapolis. Vatican News / USCCB summary reports.

Before the Facts Arrive: On the power of the first frame

  Before the Facts Arrive: On the power of the first frame Savio Saldanha SJ 22-September-2026 DOI: 10.5281/zenodo.22883874       ...