What if AI can “fake it until it makes it”?
A recent discussion at Catholic University of America raises key questions about the “humanity” of AI
One of the most important things to emerge from the recent conversation at The Catholic University of America was the back-and-forth over the nature of artificial intelligence itself. That debate matters because everything else depends on it. If we get wrong what AI is, we will get wrong how to govern it, how to use it, and how to defend the human person against its distortions.
In that sense, the conclusion reached in the discussion was both right and necessary: AI is not a being. It is not a soul. It is not a person. But I am not sure if I can agree with the description of one of the panelists; that AI is just, “a very fancy, very responsive tool.”
From a Christian anthropology, that judgment is secure. Human beings alone are created in the image and likeness of God. No artifact, however powerful, however adaptive, however eerily fluent, can cross that ontological line. A machine does not become someone merely because it imitates someone.
But that philosophical certainty should not make us comfortable.
And that, I think, was the real value of the exchange. Because if AI is “just a tool,” it is already a tool unlike any we have known before: one that speaks, persuades, flatters, advises, simulates empathy, and increasingly presents itself in forms that invite emotional and even moral confusion.
Professor Charles Camosy’s experience with top developers is especially important here. Some of the people closest to the frontier genuinely believe they may be able to move AI from “something” toward “someone.” Christians should reject that claim. But we would be foolish to dismiss the ambition behind it, or the cultural effects it may produce.
The danger is not that Christians will suddenly forget that only man bears the divine image. The danger is that a civilization trained by screens, simulations, and convenience may begin to treat what is not a person as if it were one. And that would create enormous challenges. A machine does not need actual sentience to destabilize human relationships, distort moral intuitions, weaken responsibility, or invite misplaced attachment. It need only mimic personhood persuasively enough.
That is why I hesitate to rest too comfortably in the language that AI is “only a tool,” even though the phrase is true. Hammers do not converse. Calculators do not simulate affection. Search engines do not tempt the lonely with ersatz intimacy or the confused with counterfeit wisdom. AI may remain a tool in essence while becoming, in practice, a radically new paradigm in how tools interact with human beings. That difference matters.
Yes, AI is a tool. But it may be the first tool whose manner of acting presses constantly against the boundary between instrument and apparent companion, between artifact and seeming subject. That boundary is real. Christians must defend it. But precisely because it is real, we should expect it to be contested, blurred, and exploited. The wiser course is not to assume the old categories will automatically protect us. It is to prepare now for the fact that this “fancy tool” may confront us with a genuinely new human challenge.