Over the past decade, the use of artificial intelligence (AI) has become increasingly common in many areas of science and technology. Archaeology is also beginning to explore the possibilities of AI, with dedicated sessions currently focusing on diverse applications within discipline.
Archaeology offers great benefits from AI, especially when it is necessary to analyze large amounts of data or perform complex, highly specialized, time-consuming tasks. Currently, AI is applied to many archaeological fields, including:
- Artifact Identification: AI helps categorize and date ceramic shards through image recognition.
- Site Detection: Machine learning analyzes satellite images to discover hidden archaeological sites.
- Language decoding: AI helps you decipher ancient scripts such as linear B and mucus.
- 3D Reconstruction: AI reconstructs buildings from damaged statues or fragmented artifacts. and
- Predictive Modeling: AI predicts where undiscovered artifacts are based on historical data.
AI offers enormous possibilities to promote understanding of shared archaeological heritage. However, some important issues currently need to be paid attention to. These include identifying specific archaeological research questions that are best suited for AI to address. It allows access to both the right data, its availability and the means to generate effectively. Navigating the ethical, epistemological, and interpretative challenges brought about by AI in an archaeological context. Address the lack of sustainable infrastructure and resources needed to support this task. Each of these areas is essential and requires deeper focused discussion and exploration.
Despite popular perception, one of the biggest challenges of AI is creating the datasets needed to train them, rather than developing the right algorithms. This is a problem pronounced especially in archaeology. The fields are highly digitized, but are not often converted into concrete data.
In other words, archaeological information is often stored in digital format, but is rarely structured in a way that makes it accessible or usable by machines and limits its applicability to AI. As a result, it is becoming increasingly important to ensure that archaeological data can be found, accessible, interoperable, and reusable (fair), especially in machine-accessible formats defined by fair principles.
Introducing Maia Cost Action
In September 2024, the management of Artificial Intelligence in Archaeology (MAIA) officially launched the network, bringing together digital and field archaeologists, chronology and diverse fields experts, data curators, computer scientists, museum experts, science communicators, scholars and practitioners.
This cost action aims to develop a common understanding of AI applications in archaeology, with MAIA bringing together over 255 members across 34 countries.
“Maia's cost action is a great opportunity to exchange between people in Europe and those involved in the open challenges that AI brings to every day in all fields, including archaeology. We are in a critical moment of technological acceleration in many archaeological practices. An opportunity to bring together a diverse perspective and broad expertise to tackle these challenges with fresh ideas.” Maia's Chairman, Dr. Gabrie Regattilia, Ph.D.
Maia tackles the complex and rapidly evolving relationship between archaeology and artificial intelligence. The network aims to assess current cutting edge in AI application in archaeology and to identify both its advances and limitations. It aims to create an international collaborative framework to promote the expansion of such resources in order to explore the availability and characteristics of open archaeological datasets suitable for training AI systems.
Additionally, the network focuses on understanding key archaeological research questions that will help AI address by carefully considering resource demands, ethical implications, transparency regarding data and algorithm bias, and the long-term sustainability of these technologies.
Finally, Maia strives to effectively communicate the challenges and opportunities that arise from the intersection of AI and archaeology, and share research findings and insights with a wide and diverse range of public viewers.
Kick-off in Seville: Shaping the future of AI in archaeology
In May 2025, 70 participants from over 30 countries gathered in Seville to hold a Maia kickoff meeting hosted by the University of Seville Facarto Dodegeographia Air Historia.
The Seville Conference marked the official launch of these activities and laid the foundation for Maia's future steps. Training schools and dedicated working group meetings will be launched and organized over the coming months. These provide practical opportunities to study existing archaeological datasets and AI applications around the world, especially for the creation of high-quality archaeological datasets suitable for training AI tools.
