Nov. 17, 2026, 13:30-15:00
Location: Ångström laboratory, room 106157
Registration: Link (please register by November 15)
Speaker: Adriana Iamnitchi, Maastricht University
Title: Social Media Research in the Age of AI: Data Access, Synthetic Data, and the Limits of Privacy
Abstract: Social media platforms were seen as digital environments for studying societal phenomena ranging from political polarization to coordinated information manipulation. Yet, over the past decade, independent research on these platforms has become increasingly difficult. Restrictions on data access and opaque algorithms challenge the reproducibility and reliability of computational social science. At the same time, generative AI offers a potentially transformative alternative: large language models can generate synthetic social media datasets that reproduce statistical and linguistic characteristics of real online interactions. Such datasets could facilitate reproducible research, reduce dependence on proprietary platform data, and enable the sharing of research resources. But how representative are these synthetic datasets, and how effectively do they protect the individuals whose data inspired them? In this talk, I will draw on recent empirical research to examine three interconnected challenges: the difficulties of studying social media under increasingly restricted data access; the opportunities and limitations of LLM-generated synthetic datasets; and the tension between data fidelity and individual privacy. I will discuss how authorship attribution techniques can reveal information about the creators of original content even when that content has been synthetically regenerated.
Bio: Adriana Iamnitchi is Professor and Chair of Computational Social Sciences at Maastricht University, the Netherlands. Her research spans network science, computational social science, and distributed systems, with a focus on understanding social interactions and information diffusion in large-scale online environments. Her work has investigated polarization, coordinated information campaigns, privacy, and the structural properties of social networks. More recently, she has explored the methodological challenges created by restricted access to social media data and the opportunities and risks of using generative AI to produce synthetic datasets for research. Previously, she was Professor of Computer Science and Engineering at the University of South Florida. She is an ACM Distinguished Member and recipient of an NSF CAREER Award.