AI System Just Proved The Dead Sea Scrolls Are Older Than Thought — Here’s How

GRONINGEN, Netherlands — For nearly 80 years, scholars have debated the origins of the Dead Sea Scrolls, relying on expert interpretations of ancient handwriting styles that often don’t agree. Now, a new artificial intelligence system is offering a more objective approach, with results that challenge the long-standing assumptions about when these texts were written. The findings could influence how historians understand the timeline of early Judaism and the context in which some of its most important religious writings emerged.

Researchers at the University of Groningen’s Qumran Institute developed an AI program called “Enoch” that combines radiocarbon dating with computer-based handwriting analysis to estimate the age of ancient manuscripts. When the system was applied to 135 previously undated fragments of the Dead Sea Scrolls, it consistently produced earlier dates than traditional paleographic methods. The program suggests that some scrolls may be decades older than scholars once believed.

The study, published in PLOS One, suggests that some religious texts and the movements behind them could have emerged earlier than previously thought. For example, scrolls that had been estimated to date to around 50 BCE may have actually been written as early as 150 BCE, offering new insights into the evolution of religious communities during a pivotal time in ancient Judaea.

How Scientists Trained AI to Read Ancient Handwriting

Study authors sought to improve the accuracy of manuscript dating — arguably one of archaeology’s most persistent challenges, especially when texts lack explicit historical references.

Traditional paleography — the study of ancient handwriting — can be highly subjective. Experts often arrive at different conclusions about the same manuscript, based on how they interpret shifts in letter shapes and writing styles over time.

To address this, the team developed the first AI system trained to learn from both physical evidence (radiocarbon dates) and visual handwriting patterns. They named the program “Enoch” after an ancient Jewish figure associated with wisdom and knowledge.

The researchers began by radiocarbon dating 30 manuscript samples using advanced chemical treatments that removed contaminants. Of those, 24 yielded reliable dates, which were then used to train the AI.

The system analyzed thousands of features in the handwriting, from the curvature of letters to stroke angles. These details are often invisible to the human eye but detectable by machine learning algorithms. The AI learned to associate specific handwriting features with distinct time periods using advanced statistical methods including Bayesian ridge regression.

What Did Enoch Reveal About The Dead Sea Scrolls?

When Enoch analyzed 135 previously undated scroll fragments, its predictions aligned with expert paleographic assessments 79% of the time. It’s a match rate that researchers describe as “highly unlikely to have occurred by chance alone.”

Importantly, Enoch often predicted earlier dates than traditional methods. One key example involves manuscript 4Q114, which contains text from the biblical Book of Daniel. Radiocarbon dating of this scroll yielded a range of 230-160 BCE, while scholars have traditionally dated the text based on historical content to around the 160s BCE. This demonstrates how the combined approach of radiocarbon dating and AI analysis can provide more precise age estimates.

Another significant finding concerns scrolls associated with the religious group behind the Dead Sea Scrolls. Manuscripts related to community regulations and religious practices, which had previously been dated to the first century BCE based on paleographic analysis, received earlier dates from Enoch’s AI predictions, suggesting the group’s literary activity may have begun in the second century BCE.

The AI system also revealed that two distinct handwriting styles, Hasmonaean and Herodian, likely coexisted for longer than experts once believed. This challenges the idea that one style replaced the other in a straightforward chronological sequence.

Why These Ancient Texts Still Matter

Discovered in the mid-20th century, the Dead Sea Scrolls are among the oldest surviving copies of the Hebrew Bible and provide critical context for the development of Judaism and early Christianity.

If some of these manuscripts were written decades earlier than previously believed, it could reshape scholarly understanding of how religious thought and textual traditions evolved during a crucial historical period. The results suggest that sophisticated literary and theological activity may have been flourishing in Judaea earlier than scholars realized.

The study also highlights the growing role of artificial intelligence in tackling historical questions. By combining physical data with high-resolution visual analysis, AI is helping researchers uncover patterns that are beyond human perception.

While Enoch doesn’t always agree with human experts, diverging in about 21% of cases, the authors argue that those disagreements highlight areas worthy of further investigation, rather than errors.

Looking ahead, the team plans to refine Enoch as more radiocarbon data and improved manuscript imaging become available. They’ve already tested the approach on medieval texts with known dates and achieved similar levels of accuracy, raising the possibility that AI could help redate other historical document collections around the world.

Artificial intelligence is opening a new chapter in manuscript studies, offering fresh perspectives on texts that shaped the religious and cultural foundations of the modern world.

---