The Vatican’s Apostolic Library, formally founded in the 15th century, is digitizing around 80,000 handwritten manuscripts, part of a collection that also includes 2 million books, 100,000 archival documents and hundreds of thousands of coins, medals and charts.
“People often think of the Vatican Library as a dusty old place, but in reality it has tended to be on the cutting edge,” Timothy Janz, the library’s former vice prefect and now “Scriptor Graecus,” told CNA.
To underline his point, Janz pointed to one of the many Renaissance frescoes on the walls of the Vatican Library’s Sistine Hall that depicts books stored upright on open shelves, a novelty at a time when volumes were usually placed flat.
“Being a public library was unusual in the 16th century,” he said, adding that Pope Nicholas V first described in a letter in 1451 his desire for a library “for the common convenience of scholars.”
The Vatican Library’s mission, Janz said, has always been twofold: “to make the works available to readers and also to maintain them for future readers.” Digitization, then, is “a new way of doing what the founder really wanted the library to be, to make these works available.”
The Vatican’s digitization efforts focus on its collection of unique historical manuscripts, as well as some of its oldest books, the incunabula books printed during the first period of typography before 1500.
One of the oldest manuscripts in the Vatican collection is the “Hanna Papyrus”, which is from the 3rd century AD, which has already been digitized, as has the 4th century “Codex Vaticanus”, one of the first complete manuscripts of the Bible in Greek. The digitization project began in 2012 and has so far put around 30,000 manuscripts online.
The vision is “to have a real digital library that is really usable and easy to use,” Janz said.
Elsewhere in Rome, other historic Catholic institutions are going even more high-tech.
At the Alexandria digitization center in Rome’s Historic Center, a robotic scanner turns the fragile pages of centuries-old books in the collection of the Pontifical Gregorian University Library at a speed of up to 2,500 pages per hour. In a matter of minutes, texts, some that had only been accessible to scholars traveling to Rome, can be recorded, translated and even fed into an artificial intelligence model trained to reflect Catholic teaching.
The initiative is led by Matthew Sanders, CEO of a Catholic technology firm called Longbeard, which is using robotics and AI to digitize Catholic collections at some of Rome’s historic pontifical universities and institutes.
The project began when the rector of the Pontifical Oriental Institute asked if its 200,000-volume library in the Catholic and Eastern Orthodox traditions could be accessible to scholars in the Middle East, Africa and India without requiring trips to Rome. The request was simple: digitize the books, make them readable on any device, and allow them to be translated instantly.
Since then, the Alexandria digitization center’s workload has grown. Longbeard is currently working to digitize the historical collections of the Salesian Pontifical University and the Pontifical Gregorian University and plans to work with the Pontifical University of St. Thomas Aquinas and the Venerable English College, as well as several religious orders, to digitize some or all of their collections.
Digitized works can be folded into a growing Catholic data set, training Longbeard’s AI systems such as Magisterium AI and an upcoming Catholic-specific language model, Ephrem. Institutions can choose to make their texts public or keep them private. Scholars can search collections, generate summaries, or trace an AI-generated response to its source.
The system also allows translation through Vulgate AI. Sanders recounted stumbling upon an untranslated papal document about St. Thomas More: “I never knew this existed. It was in Latin. It hadn’t been translated. We ingested it through Vulgate and suddenly I could read it.”
“When you actually go to the Hub and see a scanned book, and an hour later that work is available for anyone in the world to view in any language, that’s when you realize what this really means,” he said.
For now, the Vatican Library is taking a more cautious approach to artificial intelligence and robotics. Janz explained why he believes manuscripts in particular require a human touch rather than automation.
For scholars, he said, “The reason this manuscript is interesting is because in this specific place, it has a word that is different from other manuscripts, maybe it’s just a letter that changes it from one word to a different word,” Janz explained. “It’s that little difference that makes this book so valuable.” This type of work requires 100% precision, he added. Even if automated AI transcription reaches “99.9% accuracy… it is basically useless.”
Sanders said he “wholeheartedly” agrees that for “the deep and meticulous work of textual criticism, the original manuscript is the definitive authority, and a human expert is irreplaceable,” but added that “to limit the role of AI to mere transcription is to miss its revolutionary potential.”
“AI, even with a 99.9% accuracy rate, transforms these silent collections into a dynamic, searchable database of human knowledge,” he said. “It allows a researcher to ask, ‘Show me all the 15th-century manuscripts that discuss trade with the Ottoman Empire,’ and get instant results from collections around the world. They can identify patterns and conceptual links that were previously indescribable. The AI finds the needles in the hay; the scholar is free to perform the exact analysis on the priceless originals.”
For the Vatican Library, the digitization effort has also been integrated into its conservation efforts for these historical texts. “Every manuscript that goes into the scanners first goes to our conservation workshop and is thoroughly examined to make sure it…can withstand the strain of being digitized,” Janz said. “When the digitization is done, it goes back to the conservation workshop again, and they check to see if anything has changed.”
“We’ve discovered a lot of manuscripts that needed fixing, needed conservation work as a result of going through them all and looking at them,” he said.
Still, the Vatican Library isn’t ignoring AI entirely. He is developing a project to catalog medieval manuscript illustrations, making the images searchable by subject. In partnership with Japanese researchers, it is also training machine learning models to transcribe medieval Greek writing. “You’ll make mistakes and we tell you what the mistakes are…maybe eventually you’ll get to a point where you can do things reliably,” Janz said.
In the future, Janz said he would love to see technology make it possible to have transcriptions of all his manuscripts in historical languages available to scholars.
As for AI, he remains cautious. “I think we’re pretty open. I think we share the same concerns about AI that everyone else has.”
Inside the Sistine Hall of the Vatican Library, an ornate series of frescoes traces the long history of libraries and learning: Moses receiving the law, the Library of Alexandria, the apostles recording the Gospels. Sanders sees his AI project continuing the mission of ensuring that the wisdom of the past is “shared as widely as possible.”
“If we are to progress as a civilization, we have to learn from those who came before us,” he said. “Part of this project is making sure your reflections and ideas are available today.”
(Tagstotranslate) Vatican Library