The Second General Meeting of the MAIA COST Action (CA23141) was held in a hybrid format on March 30–31, 2026, at the Kulturfabrik Hainburg in Hainburg an der Donau, Austria.

The meeting was organized to discuss how artificial intelligence can be managed in archaeological research and integrated through interdisciplinary methods. Participants included expert researchers in archaeology, artificial intelligence, data management, and digital heritage. The event provided a significant networking platform to share the progress of current working groups, plan future projects, and develop new collaborations at the intersection of technical research and ethical frameworks.

The event began on Monday morning, March 30, 2026, with opening remarks by Action Chair Gabriele Gattiglia and local organizer Dominik Hagmann. Throughout the first day, Working Group 1 (WG1), Working Group 2 (WG2), and Working Group 3 (WG3) held simultaneous technical sessions. Within the scope of WG1, results from the literature review of 1,297 bibliographic records and the initial results of the AI mapping survey in archaeology were shared. WG2 focused on developing guidelines for training datasets across various use cases, including numismatics and image recognition.

In the afternoon workshop titled “Future Use of AI in Archaeology,” a total of 10 experts delivered “Inspiration Talks” bridging archaeology and technology perspectives:

- Cheng Liu (CNRS & Université de Bordeaux) presented a vision for a Multi-Modal Language Model for experimental archaeology.
- Ivana Pandzic (University of Banja Luka) focused on rethinking archaeology with Explainable AI (XAI).
- Marek Bundzel (Technical University of Košice) discussed how AI supports Maya archaeology.
- Konstantina Eleftheriadi (University of Cologne) presented the “Monocular Depth Estimation” method for the 3D digitization of cultural heritage artifacts.
- Muhammed Milani (Bandirma Onyedi Eylul University) explained how to understand human visual attention using Eye Tracking technology.
- Tutku Tuncali Yaman (Marmara University) discussed practical workflows, linguistic barriers, and AI-Ready processes for excavation teams.
- Michael Hedderich (LMU Munich) highlighted the importance of Large Language Models (LLM) in digitizing and analyzing archaeological archives.
- Vladimir Petrović (Serbian Academy of Sciences and Arts) shared the challenges of protecting and presenting submerged Roman remains in the Danube River.
- Carolina Collaro (University of Jaén) presented AI applications in LIDAR data analysis for archaeology.
- Patrycja Godlewska (Nicolaus Copernicus University in Toruń) discussed the use of AI as a tool for expanding the hermeneutic circle in archaeological interpretation.

The first day concluded with a social event at the Roman City of Carnuntum, an archaeological site of international importance. Participants took part in an exclusive guided tour of reconstructed Roman houses and workshops built directly on their original foundations. The evening was capped off with a curated tasting of specialities inspired by ancient Roman recipes, creating an atmospheric setting for interdisciplinary exchange.

The second day began with general feedback from the working groups and the meeting’s closing session. During the Management Committee (MC) meeting held in the afternoon, the Action’s budget status, grant planning, and leadership positions were addressed.

This meeting, held in Hainburg, is of great importance for making archaeological data AI-ready, establishing ethical standards, and disseminating scientific results.

