A groundbreaking artificial intelligence system named Aeneas is set to redefine how historians reconstruct and interpret fragments of ancient Roman history. Developed by researchers including historian Dr. Thea Sommerschield of Nottingham University and AI expert Dr. Yannis Assael from Google DeepMind, Aeneas is designed to help fill in missing words from incomplete Roman inscriptions, estimate their age, and pinpoint their geographical origins—all in moments.

The AI system draws from a database of 176,000 Roman inscriptions, many of which are damaged or only partially legible. Traditional methods of analysis have long relied on expert intuition, memory, and pattern recognition—tools which, while effective, are time-consuming and subjective. Aeneas offers a powerful upgrade: it rapidly compares incomplete texts with thousands of similar inscriptions to suggest highly probable reconstructions.

According to Professor Mary Beard of Cambridge University, a leading authority on Roman history, Aeneas could be “transformative,” not only accelerating research but also expanding the field’s capabilities. “It opens up entirely new horizons,” she said, praising the tool’s potential to identify patterns and parallels even seasoned scholars might overlook.

Ancient inscriptions, whether carved on public monuments or everyday items, are vital to understanding Roman culture, language, and governance. Yet most are incomplete, undated, or their origins are unknown. Historians typically try to reconstruct missing text using “parallels”—similar inscriptions with comparable phrases or formats. But this method is painstaking and requires long-term analysis.

“Aeneas performs this work almost instantly,” Dr. Sommerschield explained. “It’s like solving a massive jigsaw puzzle where most pieces are missing. Now, we have a tool that can scan tens of thousands of pieces and find the ones that fit.”

In one test, Aeneas was applied to the Res Gestae Divi Augusti, the monumental inscription attributed to the Roman Emperor Augustus at the Temple of Augustus in Ankara, Turkey. Its date has long been debated by historians. Aeneas narrowed down the date to two periods widely accepted by scholars: 10–20 CE and 10–1 BCE, showcasing the tool’s precision.

The system isn’t meant to replace human expertise. In trials involving 23 historians, researchers found that a collaboration between historian and machine produced more accurate results than either working alone. The AI revealed historical connections that scholars had previously missed.

“This is the real value,” said Dr. Sommerschield. “It’s not just about speeding up the process—it’s about revealing insights we hadn’t thought to look for.”

While some academics remain cautious about over-reliance on AI—especially given its tendency to misinterpret even contemporary language—developers stress that Aeneas is an aid, not a substitute. Dr. Assael clarified that its role is to suggest possibilities, not to issue final conclusions. “Historians remain the decision-makers,” he said. “AI just widens their field of vision.”

Earlier examples of AI use in classics include the digital “unwrapping” of charred Herculaneum scrolls using X-rays and deep learning. But Aeneas marks a leap forward—bridging the gap between incomplete inscriptions and meaningful historical interpretation.

As technology and scholarship converge, tools like Aeneas may not just help us understand the past more quickly, but more deeply than ever before.