Engineering Multi-agent Systems and Generative AI: Report from the Agent Toolkits 2025 Community Session

   page       BibTeX_logo.png       attach   
Andrei Ciortea, Katharine Beaumont, Gianluca Aguzzi, Matteo Baldoni, Cristina Baroglio, Amit K. Chopra, Giovanni Ciatto, Rem W. Collier, Mehdi Dastani, Angelo Ferrando, Andrea Gatti, Önder Gürcan, Timotheus Kampik, Jérémy Lemée, Somsakun Maneerat, Elisa Marengo, Viviana Mascardi, Simon Mayer, Roberto Micalizio, Guillaume Muller, Vivek Nallur, Richard Niamke, Andrei Olaru, Heloise Pajot, Chloé Petridis, I. S. W. B. Prasetya, Alessandro Ricci, Alexandru Sorici, Stefano Tedeschi, Michael Winikoff
Matteo Baldoni, Franziska Klügl, Andrei Olaru, Alexandru Sorici, Adina Magda Florea (eds.)
Multi-Agent Systems: 22nd European Conference, EUMAS 2025, Bucharest, Romania, September 3–5, 2025, Proceedings, Part I, pages 379–401
Lecture Notes in Computer Science (Lecture Notes in Artificial Intelligence) 16258
Springer
2025

Intelligent agents have been a cornerstone of Artificial Intelligence (AI) since its early days—and received significant attention in the 1990s when the notion of autonomous agent was established, giving rise to research on Engineering Multi-Agent Systems (EMAS). Traditionally, this area has focused on theories, architectures, methodologies, paradigms, and languages for designing, implementing, and governing systems of autonomous agents. More recently, advances in Generative AI—and specifically large language models—have led to a new generation of agents and multi-agent systems, often referred to as Agentic AI. However, the conceptual bridges, overlaps, and complementarities between Agentic AI and traditional EMAS research are often unclear. The Agent Toolkits 2025 community session invited positions statements from senior members of the EMAS community to discuss these recent developments. This paper summarizes and integrates their contributions and outlines key directions for research on EMAS with Generative AI.

keywordsEngineering Multi-Agent Systems, Generative AI, Agent Toolkits
origin event
container publication
page_white_acrobatMulti-Agent Systems (edited volume, 2026) — Matteo Baldoni, Franziska Klügl, Andrei Olaru, Alexandru Sorici, Adina Magda Florea