Are the Dead Sea Scrolls older than we thought? Enoch the AI thinks so

Publicly released:
International
Osama Shukir Muhammed Amin FRCP(Glasg). Creative Commons Attribution-Share Alike 4.0 International license.
Osama Shukir Muhammed Amin FRCP(Glasg). Creative Commons Attribution-Share Alike 4.0 International license.

Enoch, an AI trained on the handwriting styles of centuries-old manuscripts from the Middle East, suggests that many of the Dead Sea Scrolls might be older than traditionally thought. An international team trained Enoch on the writing style of 24 ancient manuscripts that have previously been carbon-dated, and then asked it to predict the age of 135 non-carbon-dated scrolls. They compared these predictions with traditional handwriting analysts' predictions for the Dead Sea Scrolls age, and found that Enoch's estimates for the age of the scrolls were realistic in 79% of cases, and in most cases estimated slightly older ages than the traditional analysis.

Media release

From: PLOS

AI analysis of ancient handwriting provides new age estimates for Dead Sea Scrolls

Program trained on radiocarbon dating suggests some Dead Sea Scrolls might be older than previously thought

An AI program trained to study the handwriting styles of centuries-old manuscripts from the Middle East suggests that many of the Dead Sea Scrolls might be older than previously thought, according to a study published June 4, 2025 in the open-access journal PLOS One by Mladen Popović from the University of Groningen, The Netherlands, and colleagues. This method could give researchers a new way to place undated manuscripts into the timeline of ancient history.

While some ancient manuscripts have dates written on them, giving archaeologists a precise understanding of when they were created, many manuscripts have no date information. By studying the evolution of handwriting styles over time, researchers can sometimes determine the approximate age of some undated manuscripts by evaluating their handwriting. But to use this method, researchers need enough manuscripts with accurate dates from that period of history to create a reliable timeline of handwriting styles.

For this new paper, researchers evaluated the age of historic manuscripts from various sites in modern-day Israel and the West Bank through radiocarbon dating, and then used machine learning to study the handwriting styles of each document. By pairing those two datasets together, the team could create an AI program — named Enoch, after the biblical figure — that could use the handwriting style of other manuscripts from the region to objectively determine an approximate age range.

To test the program, ancient handwriting experts evaluated Enoch’s age estimates for 135 of the Dead Sea Scrolls. The experts determined that approximately 79% of the AI’s estimates were “realistic,” with the remaining 21% determined to be either too old, too young, or indecisive.

Enoch has already helped the research team discover new things about these ancient manuscripts. For example, both Enoch and radiocarbon dating methods estimated older ages for many of the Dead Sea Scrolls than did traditional handwriting analysis. Although the authors note that more data and further research could help understand the timeline, their work provides new insights into when these documents might have been created.

The authors add: “With the Enoch tool we have opened a new door into the ancient world, like a time machine, that allows us to study the hands that wrote the Bible, especially now that we have established, for the first time, that two biblical scroll fragments come from the time of their presumed authors.”

“It is very exciting to set a significant step into solving the dating problem of the Dead Sea Scrolls and also creating a new tool that could be used to study other partially dated manuscript collections from history. This would not have been possible without the collaboration between so many different scientific disciplines, a real team effort!"

Journal/
conference:
PLOS One
Research:Paper
Organisation/s: University of Groningen, The Netherlands
Funder: This project has received funding by the European Research Council under the European Union’s Horizon 2020 research and innovation programme under grant agreement no. 640497 (HandsandBible). The funding was received by M. Popović. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Media Contact/s
Contact details are only visible to registered journalists.