
At the first Christian Art Conference, at the QEII Centre in Westminster, London, on Saturday 14 November 2026, Matthew Harvey Sanders, CEO of Longbeard, led one of the morning breakout sessions. Artificial intelligence can now supply the hand, he told the room, but it can't supply the vision. In "The Hammer, the Letter, and the Machine" he starts from the letter Pope Gregory the Great wrote to a bishop who had been smashing the images of the saints in his churches, and names three things no machine provides: makers who own every frame, sources they can trust, and people formed to see. Read the full talk below.
Cardinal Vincent Nichols, Archbishop Emeritus of Westminster, reflects on Matthew Harvey Sanders's talk at the Christian Art Conference 2026 (footage courtesy of Christian Art Today).
Last Saturday, at the Pontifical Gregorian University in Rome, I watched five short films. Every one of them had been made by people working with artificial intelligence.
One of them told the story of Jonah and the whale, and it moved me. It was so beautifully made that someone asked whether it was really all AI.
The man who made it is a Bulgarian software engineer. After a full day at his job, he'd come home and work on the film for another eight hours, sometimes into the early morning. He paid for it himself, and by his own account that was a strain. He did it anyway. The story of Jonah has been told countless times. He was convinced these new tools would let him tell it boldly, in a way no one had seen before. And he didn't make it alone. He made it with a friend, a priest studying in Rome.
Those five films were the finalists of the first CANA Awards, for Catholic Animation New Arts, and my company was glad to be one of the sponsors. It's a good name. Cana is the village where the first sign was worked on ordinary water.
Before any film came in, the organisers published five rules. The second is the one I keep coming back to. It reads: "The creator is responsible for everything in the final film: every image, voice, line, song, symbol, and edit. Do not submit raw AI output without human judgment." A few lines later it adds: "AI may assist, but the final work must be guided by a human conscience."
The rule's title is "Own every frame". Most of what I've come to say this morning is in those three words. They also answer the question about Jonah. Was it all AI? No. A machine rendered the images. An engineer and a priest made the film.
By way of background, my company was a technology and design studio before it worked in artificial intelligence. For more than ten years I've led a creative team, making video, social media artwork, logos and brands, and I still take great joy in design. So I've stood on both sides of this question: with the people who make images, and with those who build the machines that now make them.
Six weeks ago Pope Leo received a group of artists and curators at the Vatican. Echoing Paul VI in the Sistine Chapel in 1964, he told them: "the Church needs you. This is especially true in this age of artificial intelligence, in which we urgently need to distinguish human art from what is produced by machines."
I agree with him, and I'd like to offer one thought alongside his. These same machines are putting the power to make art into the hands of a great many people who never had it. The Church has every reason to welcome that, and much to offer it. His concern and this hope belong together: the further these tools spread, the more his distinction matters.

To see why this moment is new, ask who has been able to make art.
For most of history, you needed training for the hand, which meant years that someone could afford to give you. You needed materials, and time that wasn't already spoken for by the work of staying alive. And to make a film, you needed money on a scale most people will never see.
Talent was rarely the scarce thing. The means were. The engineer who made Jonah had the gift and the longing. A year ago, all the late nights in the world wouldn't have been enough.
Now think of someone with even less to spare: a woman who left school at sixteen and spent the next two decades working to feed her children. All that time she has carried a picture in her head. Say it's the Visitation as she imagines it: two pregnant women running toward each other on a hillside. She never had the years to learn to paint it. Until very recently, that picture would've died with her.
It doesn't have to now. With today's tools, she can describe what she sees and a machine will render it. She can reject what's wrong, ask again, adjust the light, change the angle of a head, and keep going until what's on the screen is what was in her mind. Once, that needed a trained hand, with a brush or with software that takes years to master.
Here's what I'll try to show you this morning. Artificial intelligence can now supply the hand. It can't supply the vision. Left to itself, it gives you the average, and an average can be lovely. But a work of art is more than lovely. It's what someone saw, and chose, and meant. That can't be averaged. Whether this becomes a flowering of Christian art or a flood of something worse depends on three things no machine provides. Makers who own every frame. Sources they can trust. And people formed to see.
I'll come to those three. But the Church has met a new power over images before, and I want to begin with how she met it.
In 597 a party of monks sent by Pope Gregory the Great landed on the Isle of Thanet, in the kingdom of Kent. Bede tells us how they came to their first audience with the king, Æthelberht: "bearing a silver cross for their banner, and the image of our Lord and Saviour painted on a board". Rome's mission to the English arrived carrying a painting.
A few years later, the same pope heard that the bishop of Marseilles, Serenus, had been smashing the images of the saints in his churches. Serenus had his reasons: he'd seen people adoring them. Gregory wrote to him, and his letters survive. He didn't fault the bishop's motive. "In that you forbade them to be adored," he wrote, "we altogether praise you; but we blame you for having broken them." Then he gave the reason the images were there at all: "For what writing presents to readers, this a picture presents to the unlearned who behold". And he gave an instruction: "if any one should wish to make images, by no means prohibit him, but by all means forbid the adoration of images."
Serenus's mistake is a tempting one. When something is being misused, breaking or banning it can feel like faithfulness. With today's tools, the real danger isn't that people will use them to make pictures. It's that they'll treat the machine as the source of what's beautiful and true: asking it what to think, what to love, what to make, and accepting whatever it hands back. That would've troubled Gregory far more than any picture.
He defended the images for the sake of the people who couldn't read, which for most of the Church's history meant most of the Church. So she taught them through what they could see: mosaics, frescoes, carved doorways, and in time the great windows.
Those windows were the work of whole workshops, glaziers and painters whose names, in most cases, we no longer know. Many hands served that message, but only ever a few, out of many millions. Ordinary believers could look at them. They could almost never make one.
Where the means were within reach, they did make things. By the 1370s the craft guilds of York were staging pageants on wagons in the streets, and by 1415 they had divided the whole of sacred history among them, from Creation to Judgment. The shipwrights gave the building of Noah's Ark, the bakers the Last Supper. We call them mystery plays now, and the name fits: in fifteenth-century York a craft was known as a "mystery", which is also the Church's word for a truth only God can reveal. For nearly two hundred years, tradesmen told the story of salvation to their neighbours with the skills they had.
A play could be made by a guild of bakers. A window, or a film, could not. That's what's changing. For the first time, the means of making even the costliest arts are coming within reach of people who could never have afforded them.

In Magnifica Humanitas, Pope Leo names three works that carry what he calls "an almost prophetic significance": Beethoven's Ninth Symphony, Picasso's Guernica, and Schindler's List. One of the three is a film.
Last month Business Insider reported on a science-fiction series, The Arc: five episodes, about eighty-three minutes in all, with scenes that move between seventeenth-century Japan and nineteenth-century Montana. A small team made it with AI video tools. Generating its images and shots cost about thirty-eight thousand dollars.
That figure leaves out the people, and the people matter. The director, Darin Stewart, said that "human ideation was top to bottom", and that everyone on the team "got paid their full day rates".
Pixar's Toy Story 5, which came out this June, reportedly cost two hundred and fifty million dollars to make. I'm not comparing the two as works of art. I'm comparing who gets to try.
Now bring that closer to home. With the Pontifical Gregorian University, my company is digitising the written patrimony of the Church, including the lives of the saints. Turning one of those lives into a film has always needed a studio and a fortune. That's one reason so few of the saints have ever reached a screen.
Imagine a teenager in Lagos telling the story of her patron in ten minutes of animation. A year ago, she couldn't have made it. This year, she can begin.
Whoever makes these films, the order has to be the same. The machine proposes. A person who knows and loves the story decides. The rest of this talk turns on that order.
Pope Leo sees both sides of this. Last December he said: "Artificial intelligence has certainly opened up new horizons for creativity". And he finished the sentence: "but it also raises serious concerns about its possible repercussions on humanity's openness to truth and beauty, and capacity for wonder and contemplation." So far I've given you the first half. Now the second, from inside the machine.
Many of today's image and video models start from pure noise, a kind of random static. Step by step they remove it, nudging what's left toward whatever best fits the words you gave them.
Pope Leo described that process in October as "statistical calculation based on millions of images created by others." Anyone who builds these systems would recognise that description, and I'd add one thing. Many of these models are tuned toward what pleases. One of the best known, Stable Diffusion, was refined in its early versions on pictures that another program predicted people would like.
I've known Emad Mostaque, who co-founded the company behind it, for many years. He has come to Rome to talk about the part the Church can play in guiding technology toward human flourishing. For a tool built for everyone, tuning it to please most people makes sense. It does mean the vision has to come from the person using it.
So when nobody is choosing, the machine gives you the most agreeable answer to the question you asked. It isn't a literal average of pixels. It's closer to the centre of gravity of everything the model has seen. Ask it for the Annunciation and it'll give you a lovely one. It won't give you Fra Angelico's, at the top of the stairs at San Marco.
Pope Leo named that danger in the same address: "In the age of technology and consumerism, the great danger is not only censorship but also homogenization". Then he made a comparison I'd love every engineer in my industry to hear. "Art is a different kind of 'search engine.' It does not proceed by indexing but by inquiry."
I make one of those engines. Magisterium AI finds what the Church has said and shows you where she said it. That's indexing, and it does it well. But it has never asked a question it wanted answered.
Three years ago Pope Francis told artists that "True beauty is always the reflection of harmony", and that harmony joins unlike things in a "unity that is not uniformity but open to multiplicity." An average works the other way. It keeps what images share and smooths away what makes each one different. One averaged picture can still please the eye. But a Church that made all her pictures by averaging would have uniformity, not harmony.
There's something deeper too. In his Letter to Artists, John Paul II reminds us that only God creates out of nothing. "The craftsman, by contrast, uses something that already exists, to which he gives form and meaning." The machine starts from what already exists too, and it can give it form. Only a person can give it meaning. John Paul wrote that artists express themselves so fully that their work becomes "a unique disclosure of their own being". A model has no inner life to disclose. It has no body to break, no failures to grieve, no one it loves. So it has nothing to put into a picture except the pictures it was shown.
Where, then, does the line run that Pope Leo asked us to draw? I don't think it runs between tools. Artists have always used them: brushes, chisels, presses, cameras. I think it runs through the maker. A film is human art when a person made its choices and answers for them. It's machine output when nobody did.
The makers of The Arc show what that looks like. They worked out the story and its settings first, with character sheets and mood boards, and only then generated the scenes. A single shot, the director said, could take eight to ten generations. Every one of those rejections was a judgment. The machine proposed. A person said no, and no again, and then yes. That refusal is where the vision lives. It's the first of the three things I named at the start: makers who own every frame.
There's a serious objection to all of this, and I want to name it plainly. Models like Stable Diffusion were trained on images gathered from across the web, many of them made by artists who were never asked and never paid. And illustrators have watched work that used to come to them go to a machine instead.
I won't argue that away, and I say it as one of the people who make these systems. Pope Leo told the artists something every company in my field should take as a duty: "Promoting education, science and culture is impossible without at the same time improving the lives of men and women who are engaged in these activities".
But the answer to a misused tool has never been Serenus's hammer. It's Gregory's letter. Keep the images. Teach people what they're for. And make sure the artists whose work taught these machines aren't the ones left behind.
With the cost of making a film falling this fast, a wave of new work is coming. First, the good news: more of the Christian story will reach more people, in more of their own languages. Some of it will be beautiful. Much of it will be careless. And some will be false in ways that look true.
Here's a failure my company spends its days guarding against. Ask a general model to write the life of a saint, and it'll often give you something fluent, moving and partly untrue. Where what it learned about the real life is thin, it tends to fill the gap with something plausible, and it seldom tells you where the gaps were. Put that script in front of a video model, and the fiction gets a face and a voice.
Imagination isn't the problem. A film that could use only a saint's recorded words would often have little but letters and prayers. The series The Chosen gives Jesus many lines that aren't in the Gospels, and it tells its viewers, up front, that dialogue has been added. The danger is invention passed off as the record. CANA's fourth rule draws that line. It begins: "AI must not be used to deceive."
That's the second of the three things: sources the new makers can trust. For the CANA Awards, we gave the creators API keys: access codes that let their own tools draw on Magisterium AI, which links each claim it makes to the passage it came from. So a filmmaker can see which lines rest on the record and which are their own, and be honest with the audience about both.
This is also why the digitising matters. A machine can only keep you faithful to sources it can read. In Rome, at the Alexandria Digitization Hub, the books of the pontifical universities' libraries are being scanned on machines built to handle fragile volumes without harming them. The originals stay on their shelves, and get to rest. What changes is that anyone, anywhere, can find what's in them.
Accuracy isn't the only thing at stake. So is reverence. CANA's third rule puts it this way: "Sacred things should not be used to impress; they should help the viewer encounter truth and beauty." A machine renders the face of Christ by the same process it uses for any other request. Nothing in that process sets this face apart. Reverence has to come from the person directing it.
Pope Leo, borrowing John Paul II's words, asks whether this technology makes life "more worthy of man". A ten-minute animation of a saint's life, faithful to the sources, made by a teenager who has never held a paintbrush: I think that passes. A thousand cheap clips of a weeping statue, generated as bait for clicks: that fails, and the CANA rules forbid it in so many words.
Their first rule needs only seven: "Let the Gospel lead, not the machine".

Underneath all of this is a larger change. Machines are taking on more and more of the work people do for a living, at desks first and increasingly with their hands. That will bring hardship, and I don't make light of it.
But it also means that, within our own working lives, a great many people may have hours that no employer needs. If the gains of these machines are shared, and that's a real if, far more of them will get the chance to find out what they were made for. I believe many will discover they were made to make things.
Still, hours don't make artists on their own. Free time without formation tends to make consumers, not craftsmen, and that's true of most of us. Something will gladly fill those hours for us: a feed that never ends. That's why the third thing matters most: people formed to see.
Here's the distinction I most want to leave with you. There are two kinds of difficulty in making something. One stands between you and the work. The other is the work itself.
Take the first. In Chaucer's day, the painter Cennino Cennini wrote that an apprentice should spend "a good six years" grinding colours, boiling glue and preparing panels before he began to paint. For a filmmaker today, that wall is the camera, the crew and the money. That's the difficulty these machines are taking away.
The second kind is different, and you know it far better than I do: learning to see, deciding what to leave out, finding the one gesture that tells the truth about a face. No machine can lift that from you, and none should. That difficulty is what forms the person who does it. Hand it over too, and you're back to the average.
The woman with the Visitation in her head will find this out. The machine can give her that picture in minutes. Knowing whether it's true may take her years, and those years are what will make her an artist.
That's what I mean by the return of the artisan. The men of the York guilds were tradesmen, formed by their craft, making something for the people they lived among. And the long years Cennini describes were about more than colours. You could study his book by day and by night, he told his readers, but without practice under a master "you will never amount to anything". The machine can grind the colours now. It can't take the master's place.
The Church has been the patron of artists for most of her history. She could now be the patron of thousands of new makers at once, and each of the three things I've described suggests a way to begin.
The first is to find the makers, and to hold them to the standard of owning every frame. A competition, a screening night, a commission for a feast day: each one tells them the Church wants what they can make, and what she expects of it.
The second is to open the sources. A diocese that digitises its records, and shares the insight of its scholars, gives every future filmmaker something true to work from. That's work my company does, but it isn't work any one company should own.
The third is formation, and it falls to the people in this room. Many of you have spent your lives learning exactly what the new makers lack. The teenager in Lagos with her saint's story needs someone to tell her the light is wrong, the face is lying, and the scene says something the saint never meant. No model can do that for her. You can. May I ask something of you? Take an apprentice. Judge a competition. Give an evening to the young makers in your parish. The machine can make them faster. Only you can make them better.
And the people who watch these films will need something from the rest of us. A film can make someone want to know a saint. It can't introduce them to one. For that they need people who already know that saint, through years of prayer and imitation, and who'll walk with them into a church. The new makers can open the door. Behind it there has to be a community.
What I hope for isn't a flood of content. It's small workshops everywhere: in parish halls and school art rooms, and at kitchen tables. I've already seen one: a Bulgarian engineer and a priest in Rome, making Jonah. Now picture the rest. A teenager in Lagos bringing her patron's life to the screen, faithful to the record and honest about what she's imagined. A woman who left school at sixteen, seeing her Visitation at last. And beside each of them, one of you, teaching them to see.
For many people now, a screen may be where they first meet the story of Christ. Let what they meet there be true, and let it be beautiful. Then the screen can do what the painted board did in Kent in 597, and carry his face to people who have never seen it.
That face is why Christians make images at all. Before the Incarnation, as Saint John of Damascus wrote, God "could not be represented by an image". Then God became man. He didn't become an average. He became one man, with one face, from one small village in Galilee. And Saint Paul could write of the Son of God, "who loved me and gave himself for me." Every one of us can say the same. No one has ever been loved on average.
That's why his face can't be left to a machine alone. It can supply the hand. The vision belongs to those who know that love.
Every image of him, painted on a board, set in glass, or rendered by a machine, is only a glimpse. Help the new makers keep that glimpse true. For now, as the same apostle wrote, we see "in a mirror dimly". But he told us how it ends: "then face to face."
Thank you.