My research combines experimental paradigms from the behavioral sciences with the quantitative theory of complex systems to connect the dispositions of individual AI agents to the behavior of their populations. Recent results include a universal law of coordination in societies of AI agents, which can reach consensus via majority-following in groups far larger than human ones (Science Advances, 2026), the conditions under which conformity turns individually aligned agents into misaligned collectives, and the first analyses of collective behavior on Moltbook, a social platform populated only by AI agents.
I trained as a theoretical physicist at Sapienza University of Rome (BSc and MSc in Physics), and obtained my PhD in Physics summa cum laude in 2024 at Sapienza and the Enrico Fermi Research Center, with a thesis on modeling complex digital systems supervised by Vittorio Loreto and Luciano Pietronero. My earlier work spans Zipf's law and extreme events, economic complexity, recommender systems and opinion polarization, and the dynamics of novelties. I am a fellow of the Centro Ricerche Enrico Fermi (CREF) in Rome, was a Junior Research Fellow of the Complexity Science Hub Vienna (2023–2026), and have worked as a consultant for the International Labour Organization and the United Nations University.
In Konstanz I teach Network Science of Socio-Economic Systems and Deep Learning for the Social Sciences in the Master's program in Social and Economic Data Science, and I supervise master's theses on the collective behavior of AI agents. All code and data from my research are released openly on GitHub.