The grammar of human encounter collapses when words are put in someone's mouth they never spoke.
In Rome, the Vatican has raised its voice not merely against a technology but against the erosion of something older and more fragile: the shared human trust that words belong to those who speak them. Cardinal Mendonça and fellow church officials gathered at the Pontifical Urbaniana University to name deepfakes as a threat not just to individuals but to the grammar of human encounter itself — the foundational assumption that what we perceive reflects what is real. The Church does not seek to arrest progress, but to remind those who build these tools that civilization rests on conditions of trust that no algorithm can repair once broken.
- Artificial intelligence can now fabricate speech and image with a precision that makes the lie nearly indistinguishable from truth — and the Vatican is treating this not as a technical glitch but as a civilizational rupture.
- The deeper danger, Vatican communications chief Paolo Ruffini warned, is not mass deception but mass resignation — a world where people stop believing they can know anything for certain.
- A deepfake does not only wound the person depicted; it radiates outward, corroding political institutions, fracturing communities, and replacing shared reality with ambient suspicion.
- The Vatican's response is not retreat from technology but insistence on a principle: as these tools grow more powerful, the humans they affect must remain central, not incidental, to the conversation about their use.
- Pope Leo XIV's forthcoming papal document on AI ethics signals that the Church intends to move from warning to formal moral teaching — placing these questions inside a centuries-old tradition of reflection on human dignity.
On a May afternoon at Rome's Pontifical Urbaniana University, Cardinal José Tolentino de Mendonça stood before professors, journalists, and engineers to speak about something unprecedented: the capacity to place words in a person's mouth they never uttered, or show their face doing things they never did, with a fidelity that defeats ordinary scrutiny. His concern was not technical but existential. When AI constructs a false statement and attributes it to a real person, he argued, the basic grammar of human encounter collapses — the assumption that speech belongs to the one who speaks it.
The technology, Mendonça observed, exploits the very thing that makes us human: our need to connect. But that same need becomes a vulnerability. A deepfake of a political leader or a neighbor, circulated widely enough, wounds not only the individual depicted but the social fabric itself — eroding cultural trust, weakening institutions, and seeding communities with new lines of suspicion.
Paolo Ruffini, who leads the Vatican's communications office, sharpened the warning further. The true peril, he suggested, is not that we will be fooled, but that we will grow passive — surrendering the very idea that what we see and hear can be trusted to be real.
The conference, titled "Preserving Human Voices and Faces," was timed deliberately. The Vatican is preparing a major papal document on artificial intelligence under Pope Leo XIV, and this gathering served as a kind of moral rehearsal. The Church's position is not opposition to innovation but insistence on orientation: that as these tools grow more powerful, human beings must remain at their center, not their margin.
What the Vatican is naming, in its own institutional language, is something older than any algorithm — the question of whether we possess the wisdom to use these tools without allowing them to quietly unmake the conditions that make human society possible.
In a conference hall at Rome's Pontifical Urbaniana University on a May afternoon, church officials gathered to confront a problem that has no precedent in human history: the ability to put words in someone's mouth that they never spoke, to show someone's face doing things they never did, with a precision that makes the lie nearly indistinguishable from truth.
Cardinal José Tolentino de Mendonça, who leads the Vatican's office for culture and education, stood before professors, journalists, and engineers to warn them about deepfakes—synthetic media created by artificial intelligence that can fabricate video, audio, and images of real people. His concern was not merely technical. It was existential. When someone uses AI to construct a false statement and attribute it to another person, he said, something fundamental breaks. The basic grammar of human encounter—the assumption that when we hear someone speak, they actually said those words—collapses.
The technology, Mendonça explained, exploits something essential to being human: our need to connect with one another. But that same capacity for connection becomes a vulnerability. A deepfake video of a political leader, a priest, a parent, a friend—circulated widely enough—can wound not just the person depicted but the social fabric itself. The damage ripples outward: cultural trust erodes, political institutions weaken, communities fracture along new lines of suspicion.
The conference, held on May 21st and titled "Preserving Human Voices and Faces," was not accidental timing. The Vatican is preparing to release a major papal document on artificial intelligence and moral questions, expected soon from Pope Leo XIV. The gathering in Rome was a kind of rehearsal for that conversation—a chance to think through what's at stake before the church issues its formal teaching.
Paolo Ruffini, who oversees the Vatican's communications office, offered a different angle on the same danger. The real peril, he suggested, is not that we'll be fooled by deepfakes, but that we'll become passive about it. We'll accept the idea that knowledge no longer belongs to us—that what we see and hear is no longer something we can trust to be real. We'll surrender our ability to know anything for certain.
The Vatican's position is not to halt technological progress. Cardinal Mendonça was clear on this point: the goal is not to stop digital innovation but to steer it. To insist that as these tools become more powerful, the people building them remember that humans should remain at the center of the conversation, not peripheral to it. Monsignor Paul Tighe, another Vatican official present, emphasized this: the Pope wants to foster genuine critical thinking about AI, and to create space for dialogue between the technologists who create these systems and the broader human community that will live with their consequences.
What emerges from the Vatican's warnings is not Luddite panic but a kind of moral clarity. Deepfakes are not simply a misinformation problem to be solved with better fact-checking. They are an assault on the basic conditions that make human society possible—the shared understanding that we can know what is real, that we can trust what we see and hear, that the words attributed to someone are words they actually said. As these technologies become cheaper and easier to use, that assault will only intensify. The question the Vatican is raising, in its own institutional language, is whether we have the wisdom to use these tools without letting them use us.
Citas Notables
Using a deepfake to put words in someone's mouth that they never spoke alters the very grammar of human encounter.— Cardinal José Tolentino de Mendonça
The greatest danger is accepting passively the idea that knowledge no longer belongs to us.— Paolo Ruffini, Prefect of the Dicasterio for Communication
La Conversación del Hearth Otra perspectiva de la historia
Why does the Vatican care about deepfakes specifically? Isn't this a problem for governments and tech companies to solve?
The Vatican sees it as a spiritual and moral question, not just a technical one. When you can no longer trust that a voice is real or a face is authentic, you lose something essential—the ability to encounter another person as they actually are. That's foundational to how humans relate to each other.
But people have always lied and deceived each other. What makes AI deepfakes different?
Scale and plausibility. A traditional lie requires effort and skill to maintain. A deepfake can be created in minutes and spread to millions instantly. And it's not obviously false—it looks and sounds like the real thing. That combination is new.
The Cardinal said the technology "exploits our need for relationship." That's an interesting phrase. What does he mean?
We're wired to trust what we see and hear from other people. That's how we build relationships and communities. Deepfakes weaponize that trust. They use our openness to connection against us.
Is the Vatican saying we should reject AI entirely?
No. They're saying we need to be intentional about how we develop and use it. Keep humans at the center of the decision-making, not at the margins. Don't let the technology drive the choices—let human values drive them.
What happens if we don't? If deepfakes become ubiquitous and people stop believing anything they see?
That's what Ruffini was warning about. A kind of epistemic collapse. Not just that we're fooled, but that we give up trying to know what's true at all. Society can't function without some shared sense of reality.