Magnifica Humanitas on Deepfakes

The Pope on the receiving end

Begin with what is unusual about this encyclical's author. Pope Leo XIV is the first pope to write a major document on AI as someone who has himself been the repeated target of AI-generated impersonation.

Since his election in May 2025, dozens of deepfake videos have circulated showing the Pope endorsing political candidates he has never met, condemning groups he has never criticized, blessing products he has never seen, and saying things he has never said. The videos have been convincing enough to fool ordinary viewers and have been amplified by social media systems that do not distinguish between real and synthetic at the point of distribution. The Vatican Dicastery for Communication issued a formal warning in September 2025 acknowledging the problem and asking platforms and the public to be vigilant.

This is the immediate context for Pope Leo XIV's recurring insistence that faces and voices are sacred. It is the language of someone who has watched his own face used against him. The 2026 World Communications Day message, titled Preserving Human Voices and Faces, was published May 17, 2026, eight days before the encyclical it foreshadowed. The encyclical extends the argument.

It matters that the argument is being made by someone with skin in it. The encyclical is not a detached academic intervention. It is the considered theological response of a person who has been deepfaked, addressed to a world in which everyone, in time, will be.

The trajectory: how the Church arrived at this argument

The encyclical's argument on deepfakes did not emerge in May 2026. It has been building for years. The key moments, briefly.

February 2020. The Rome Call for AI Ethics identified transparency and impartiality as core principles for AI systems. The principles were stated at a high level and did not yet name synthetic media specifically, but the framework was being laid.

January 2024. Pope Francis's World Day of Peace message on AI warned that the technology was being used to generate misinformation and to manipulate public opinion. The 2024 message was the first major papal statement to take AI-generated content as a specific moral problem.

June 2024. Pope Francis addressed the G7 summit at Borgo Egnazia. He warned that AI risked turning human relations into algorithms and called for human control over decisions of consequence. The G7 address did not focus on deepfakes but did establish that AI's mediation of human encounter was a Vatican concern at the highest level.

January 2025. The Vatican Dicastery for the Doctrine of the Faith released Antiqua et Nova. The doctrinal note addressed AI deception across multiple domains and called the use of AI to deceive in education or in relationships, including human sexuality, "immoral" and requiring "careful vigilance" (n. 70). The framework that would inform the encyclical was now established at the level of formal Vatican teaching.

May 2025. Pope Francis dies. Cardinal Robert Prevost is elected and takes the name Leo XIV.

Summer 2025. Deepfake videos of Pope Leo XIV begin circulating. The volume increases through the summer.

September 2025. The Vatican Dicastery for Communication issues a formal warning on AI-generated content depicting the Pope and other Church figures. The warning is, as far as I can determine, the first such Vatican statement on synthetic media of a specific identifiable person.

May 17, 2026. Pope Leo XIV releases the World Communications Day message Preserving Human Voices and Faces. The phrase "faces and voices are sacred" appears here in its most direct form. The message is, in retrospect, the public preview of the encyclical.

May 25, 2026. Magnifica Humanitas releases. The argument the Church has been building since 2020 reaches its magisterial form.

The point of the timeline is that the encyclical is not a reaction to a moment. It is the consolidation of a trajectory.

The central argument: why faces and voices are sacred

The encyclical's argument on deepfakes runs through a specific theological claim that deserves to be stated carefully, because it is the move that does the work.

The starting point is the Catholic understanding of the human person as imago Dei, made in the image of God. This is not a vague gesture toward human worth. It is a specific claim about what a person is: a being whose dignity is intrinsic, irreducible, and reflects something of God's own being.

The face and voice, in the Catholic tradition, are not arbitrary surface features of the person. They are how the person is present to others. The face is where one human being recognizes another as a person and not a thing. The voice is how persons address one another, make promises, ask forgiveness, declare love, and bear witness. The whole structure of human communion is built on the encounter between faces and the exchange of voices.

This is why Pope Leo XIV's 2026 World Communications Day message describes faces and voices as something God gave us "when he called us to life through the Word he addressed to us." The framing is theological. The face and voice are not merely biological. They are the medium of the address through which God constitutes the human person as a being capable of being called and capable of responding.

From that starting point, the argument on deepfakes follows.

A deepfake generates a counterfeit face and a counterfeit voice. It uses these counterfeits to put words in a person's mouth that the person never spoke, to associate the person with positions they never held, to make them appear in scenes they were never in. The deepfake is not, in the Catholic frame, a problem of false information that happens to use someone's likeness. It is a problem of false personhood. The image of the person has been used against the person. The face that was given for communion has been weaponized for deception.

This is why the encyclical's objection is theological rather than purely technological. A technical objection would focus on detection, watermarking, platform liability, all of which matter and all of which the encyclical may well address. But the deeper objection is that something is being done to the person that the Catholic tradition has never had words for, because the technology that does it is new, and the framework for addressing it has to be built. Magnifica Humanitas is the first sustained Catholic attempt to build that framework.

The anchor quotes that frame the encyclical

Three statements by Pope Leo XIV, drawn from the eight months before the encyclical, give the texture of what readers should expect.

"Faces and voices are sacred. God, who created us in his image and likeness, gave them to us when he called us to life through the Word he addressed to us."

Pope Leo XIV, 2026 World Communications Day Message (May 17, 2026)

"By simulating human voices, faces, emotions, and relationships, the systems known as artificial intelligence not only interfere with information ecosystems, but also encroach upon the deepest level of communication, that of human relationships. The challenge, therefore, is not technological, but anthropological. Safeguarding faces and voices ultimately means safeguarding ourselves."

Pope Leo XIV, 2026 World Communications Day Message (May 17, 2026)

"We need faces and voices to speak for people again."

Pope Leo XIV, address to journalists, 2026

The encyclical is expected to extend each of these into a developed argument. The first becomes the theological anchor for the entire treatment of synthetic media. The second becomes the bridge from individual harm (a person whose face has been counterfeited) to systemic harm (a public sphere in which the very category of authentic representation has been destabilized). The third becomes the practical exhortation: the response to synthetic substitutes is not less communication but more authentic communication.

What the encyclical likely says, and what to look for in the text

The full text releases May 25 and this section will be revised then. Based on the trajectory above and on Pope Leo XIV's prior statements, the encyclical is expected to develop the argument on synthetic media in roughly four moves.

First, a theological statement of why faces and voices are sacred, grounded in the imago Dei and in the Catholic understanding of human communion. This is the foundation. Watch for citations of Genesis 1, Gaudium et Spes, and the patristic tradition on the face as the site of personal encounter.

Second, a description of the harm that synthetic media inflicts, framed not just as misinformation but as a violation of the personal integrity of the depicted. Watch for language about deception, calumny, and false witness, all of which are categories the Catholic moral tradition has developed at length.

Third, a particular concern for the most vulnerable: minors targeted by AI-generated intimate imagery, women targeted by non-consensual synthetic content, public figures targeted by political deepfakes, and ordinary people whose images are scraped for training and reproduction without their consent. Watch for explicit naming of these categories.

Fourth, a call to public authorities, AI developers, platforms, and individual users. The Catholic tradition does not typically prescribe the precise legal mechanisms, but it does identify the moral problem and call for action. Watch for the language addressing each of these groups by name.

If the encyclical does what it is expected to do, the section on synthetic media will be one of its most quoted and most politically consequential. The EU AI Act enters its main enforcement phase in August 2026. Synthetic media governance is being debated in legislatures around the world. The encyclical's framing will enter that debate immediately.

Why the theme runs through the whole encyclical

A reader who comes to Magnifica Humanitas looking only for the deepfakes section will miss what the encyclical is actually doing. The argument about synthetic faces and voices is not confined to one chapter. It is the model the encyclical uses to think about every form of AI-mediated simulation, and the language returns throughout the document.

AI companions that simulate friendship use synthetic faces and voices to do so. The encyclical's argument that this is a violation of the person depicted extends, with slightly different force, to the argument that AI companions for the lonely are a violation of what genuine companionship is. The same framework that condemns deepfakes also informs the encyclical's treatment of AI in care relationships and in the formation of children.

AI-generated content that mimics human creativity, where the model is trained on the work of countless human artists and produces output that imitates without acknowledging, is treated by the encyclical in similar terms. The face of the artist has been used without consent. The voice of the writer has been ventriloquized without permission. The pattern is the same.

AI systems that generate misinformation, manipulate elections, or shape political opinion through synthetic content are addressed under the same anthropological frame. The argument is that all of these uses of AI share a common structure: they substitute fabricated representations of persons for the persons themselves, and they do so at a scale and with a verisimilitude that the older categories of deception and slander were never built to address.

This is why "faces and voices are sacred" is not a slogan in the encyclical. It is a principle that does work across the entire document.

What this means for ordinary readers

For Catholic readers, the encyclical is, among other things, an examination of conscience. The question is not only what governments and platforms should do. It is what ordinary Catholics participate in when they share, consume, or create synthetic content. Forwarding a deepfake video, even with the intention of mocking it, contributes to its reach. Using AI tools to generate images of real people for private amusement participates in the pattern the encyclical is naming. The encyclical is likely to be specific about this, and the specificity will be uncomfortable.

For journalists and policymakers, the encyclical offers a framework that is not currently well-represented in the synthetic media governance debate. The standard secular framings are about misinformation, electoral integrity, and consumer protection. The Catholic framing adds something the secular debate has been missing: a theologically grounded account of why the harm of synthetic media is a harm to the person depicted, not merely a harm to the information ecosystem. That account will be useful to anyone trying to make the case for stronger protections.

For AI developers and platforms, the encyclical is likely to read as an appeal to conscience that does not pretend the issue can be resolved by technical means alone. Detection systems matter and watermarking matters, but neither will undo the moral wrong of producing or hosting content that puts words in a real person's mouth without their consent. The encyclical's call will be to refuse to build the systems that make this trivial, not merely to label the outputs.

For everyone, the encyclical asks a question that the trajectory of the technology has made unavoidable: in a world in which any face can be generated and any voice can be synthesized, what does it mean to encounter a real person? The Catholic answer is that the encounter is sacred, that the encounter is what we are made for, and that the technology that obscures it has to be ordered to the human good rather than the other way around.