When I first read the Pledge Against AI Catholic Content and Text (PAACCT), published by Catholic Culture on October 7, I found myself in an uncomfortable position: I agreed with much of its diagnosis, sympathized with its intentions, and nevertheless reached a very clear conclusion: I will not sign it.
Before explaining why, let me make something absolutely clear. Among the signatories are Catholic writers, intellectuals, and journalists whom I deeply respect, including some whom I consider good friends. I have no doubt about their commitment to the Catholic faith and their legitimate concerns about the profound consequences of artificial intelligence for human creativity, communication, and ultimately human dignity.
In fact, those concerns are precisely why I launched this Substack.
From its beginning, Imago Dei has been an attempt to confront the extraordinary technological revolution unfolding before our eyes through the lens of Catholic anthropology. I have repeatedly argued that the consequences of AI cannot be reduced to questions of productivity, technological efficiency, economic growth, or even the potentially devastating displacement of human labor.
I have written about the disturbing proliferation of AI-generated Catholic books and content, about the threat to authentic Catholic writing so perceptively identified by Amy Welborn -one of the signatories-, about the disruption of employment and the dignity of work, and about the temptation to substitute artificial companionship for genuine human relationships.
I have never subscribed to the naive technological optimism that assumes every innovation must necessarily represent progress.
But neither am I prepared to embrace the opposite assumption: that because a technology presents unprecedented dangers, the appropriate Catholic response is to draw a series of prohibitive lines around its use while we are still struggling to understand what it is becoming. And that is where my disagreement with PAACCT begins.
Sound anthropology, questionable prohibitions
The pledge opens by invoking Pope Leo XIV’s encyclical Magnifica Humanitas, particularly its distinction between human intelligence and artificial intelligence.
On this fundamental point, I have no disagreement.
Human intelligence is not merely the capacity to manipulate information, recognize patterns, or produce persuasive language. The human person is an embodied, relational, moral, and spiritual being, capable of knowing truth, exercising freedom, assuming responsibility, and entering into communion with God and others.
AI does not become a person by imitating human language, producing an apparently original argument, or generating an extraordinarily beautiful image. The distinction is fundamental, and Catholics should defend it without hesitation.
But accepting the anthropological premises of PAACCT does not logically require accepting the particular prohibitions its authors derive from them. Some of these provisions are sensible professional standards that should already govern responsible journalism and publishing, regardless of whether AI is involved.
But the package as a whole strikes me as a collection of prudential judgments elevated into something resembling a definitive Catholic response.
With the acceptance of some use of AI, the pledge recognizes that AI is not intrinsically objectionable and that its moral significance depends, at least in part, on how it is used.
Yet the boundaries they establish between acceptable and unacceptable uses often appear arbitrary:
Why should using AI to translate a text be permissible, while using it to improve the expression of an author’s original argument should be prohibited? Why may AI help organize research or stimulate ideas, but not assist substantially in organizing those ideas into written form? And why is generating a religious image necessarily an act of simulated worship rather than, in some circumstances, the production of a visual instrument whose religious significance depends on human intention and use?
I pose these as honest questions, not as an attempt to dismiss the moral issues involved. I in fact agree with the pledge’s invocation of St. John Henry Newman concerning the intimate relationship between thought and language; but that does not establish that every form of technological assistance in the process of expression necessarily diminishes intellectual authorship.
As someone who needs editing in the way I express myself in a language that is not my native, I believe that an author who uses AI to polish the English expression of an argument he has conceived, developed, examined, and ultimately approved has not necessarily surrendered his intellect or creativity.
Needless to say,I am convinced that presenting machine-generated ideas as one’s own, fabricating sources, disguising plagiarism, or abandoning intellectual responsibility are serious moral and professional failures. But I am equally convinced that we cannot simply identify the involvement of generative AI with the abdication of human thought.
Immutable anthropology, rapidly changing technology
My second objection is based on the fact that the Catholic understanding of the human person is not subject to technological revision. The dignity of the human person does not increase or diminish according to the sophistication of the machines we develop.
But the technology we are evaluating is not immutable. Quite the opposite: AI is evolving at an astonishing pace, and we are still in the early stages of understanding its possibilities and limitations.
In previous posts, I have discussed the work of Dean Ball and others who are exploring one of the most fascinating questions in AI development: interpretability.
We are increasingly capable of creating systems that produce sophisticated results without possessing a complete understanding of the processes through which those results emerge.
That should be a source of considerable concern. But it should also inspire intellectual humility.
We do not yet know how far these technologies will develop, which present limitations will remain permanent, or how increasingly sophisticated AI systems might interact with human reasoning, learning, and creativity.
Most importantly, we do not know whether future forms of human-AI interaction might actually expand certain capacities of human intelligence rather than merely replacing them.
In short, our understanding of how machines can assist human intelligence remains incomplete. This is precisely why I resist committing myself to permanent-sounding prohibitions based on the present state of the technology.
Lessons from a Benedictine abbot and the printing press
There is a remarkable historical precedent that I believe deserves serious consideration by those who have signed PAACCT, as well as those contemplating doing so.
It involves the great Benedictine scholar Johannes Trithemius, abbot of Sponheim, who lived during the revolutionary transition from manuscript copying to printing.
In his famous De laude scriptorum manualium (In Praise of Scribes), written in the final decade of the fifteenth century, Trithemius defended the spiritual, intellectual, and cultural importance of the monastic tradition of copying manuscripts by hand.
Trithemius was probably the brightest Catholic mind of his time, and as such, he was convinced that the work of the monastic scribes was not merely an efficient means of reproducing texts. It was a spiritual discipline, an exercise in patience, contemplation, learning, and devotion. The copying of sacred texts was deeply integrated into the religious life of the monastery.
Trithemius was worried that the printing press threatened to undermine an entire culture of intellectual and spiritual formation. He was also concerned about the inferior quality of printed books, their potential fragility, and the disappearance of the meticulous craftsmanship that characterized the finest manuscripts.
He was absolutely right about every single concern and criticism; but history had other plans. Here I strongly recommend an extraordinary essay by the scholar James J. O’Donnell, The Pragmatics of the New: Trithemius, McLuhan, Cassiodorus.
O’Donnell’s analysis is particularly illuminating because he does not caricature Trithemius as a reactionary incapable of recognizing technological progress. Indeed, Trithemius himself made use of printing and eventually praised it.
O’Donnell’s central insight is that many of the early criticisms of printing were substantially correct. Printed texts could contain errors, spread misinformation, undermine established intellectual practices, and disrupt institutions that had served society for centuries.
Yet those criticisms failed to anticipate how the new technology would also create opportunities to address its defects and develop possibilities that were almost impossible to imagine at the beginning. And the Catholic Church eventually became one of its great beneficiaries.
I am aware that the historical analogy is not perfect. AI presents questions concerning human cognition and creativity that go well beyond those raised by printing.
But I think that the main point stands: a criticism can be theologically sound, morally serious, and factually accurate… and still lead to a mistaken practical judgment about how we should respond to a new technology.
This is the danger of drawing what I call the last battle line too early. Catholics, instead, should be especially careful about allowing a legitimate defense of permanent truths to become an unnecessary defense of historically contingent ways of doing things.
The problem with treating technological restraint as a Catholic identity
PAACCT’s authors explicitly clarify that it is neither a vow nor an oath and does not, by itself, bind signatories under pain of sin. Good!
Nevertheless, a public pledge endorsed by respected Catholic intellectuals, media organizations, and cultural institutions inevitably acquires significance beyond the personal commitments of its signatories. Thus, it can gradually become a marker of Catholic credibility.
And that creates the risk of suggesting, even unintentionally, that those who embrace a different prudential judgment about AI are somehow less committed to authentic Catholic anthropology or the defense of human creativity.
I reject that implication.
There is no reason why a Catholic who chooses not to use generative AI in his writing should be regarded as intellectually backward. Nor is there any reason why a Catholic who carefully uses AI as an instrument of his own intellectual activity should automatically be suspected of compromising human dignity.
Prudence is not a synonym for prohibition
This brings me to what I consider the most important point of all: the Catholic virtue of prudence. Prudence, in its classical Thomistic understanding, is not simply caution, hesitation, or the avoidance of risk. Aquinas describes it as recta ratio agibilium; right reason applied to action.
It is the virtue that allows us to discern the appropriate means of pursuing genuine goods under concrete and often uncertain circumstances. It is not the virtue of standing still, nor an excuse for surrendering to technological inevitability.
Authentic prudence sometimes demands a firm refusal. There are uses of AI that already deserve categorical moral rejection. But prudence also recognizes the difference between an intrinsically wrong action and a legitimate human activity that can be carried out well or badly depending on intention, means, and circumstances.
Most importantly, prudence resists the temptation to offer definitive solutions to problems whose dimensions we do not yet fully understand.
I have repeatedly argued in this Substack that Catholics should neither surrender the future of artificial intelligence to the technological industry nor retreat into a defensive posture that leaves us increasingly irrelevant to its development.
But we need to avoid repeating an old historical mistake: defending an eternal truth by fastening it unnecessarily to a temporary technological arrangement.
Why my answer remains no
I understand the desire of PAACCT’s signatories to make a public stand against the increasingly indiscriminate use of artificial intelligence. In many respects, I share their frustration.
But my answer to these problems is not to sign a document whose specific prohibitions I cannot adequately justify, either from its theological premises or from what we currently understand about the technology.
I prefer a path that requires continuous moral discernment instead of a fixed list of acceptable and unacceptable technological uses.
Of course, from my small corner in the Catholic church, I will continue to defend human dignity against the dangers of AI. I will continue to criticize its abuses, challenge its most extravagant promises, and question the economic and cultural forces driving its development.
But I will not draw the last battle line while the landscape is still changing beneath our feet.


