EC2 Supports Groundbreaking Research on AI and Filter Bubbles

How do opinions really form in today’s digital societies? And why do online discussions so often fragment into polarised communities, commonly described as filter bubbles?

Within the EC2 project, we are proud to support innovative research that addresses these questions from a fresh and methodologically ambitious perspective. A newly published study investigates how filter bubbles can emerge when people exchange ideas expressed in natural language, rather than simplified numerical opinions.

By combining agent-based modelling with large language models, the research moves beyond abstract representations of opinion and instead captures how real-world beliefs are articulated, interpreted, and transformed through communication. This approach opens new ways of understanding how worldviews evolve, spread, and cluster in complex digital environments.

Why this research matters for EC2

This work reflects exactly what EC2 stands for: pushing the boundaries of social science by integrating advanced computational methods, artificial intelligence, and democratic theory. Instead of treating opinions as static or isolated, the study models them as dynamic, language-based constructs shaped by interaction and context.

Thanks to EC2, researchers can explore bold ideas, test novel modelling approaches, and better capture the complexity of social interaction in the digital age. Such experiments help illuminate not only how polarisation emerges, but also how democratic discourse might be better understood — and ultimately strengthened — in increasingly mediated societies.

From AI models to democratic insight

Rather than offering simple answers or technical fixes, the research supported by EC2 provides analytical tools for reflection. By simulating how language, cognition, and interaction intertwine, it helps clarify the mechanisms behind information clustering and belief formation.

These insights are essential for anyone interested in democracy, public discourse, and the societal impacts of AI — from researchers and policymakers to practitioners working on digital governance and platform design.