Executive Summary

This report presents key insights from the 1st Symposium on Diversity of News Media Diets and its Effect on Democratic Citizenship held in April 2025 in Amsterdam. The event  gathered 35 experts in the field and panellists from six European countries. It was convened thanks to the financial support of the Royal Netherlands Academy of Science (KNAW) Early Career Partnership program awarded to Sílvia Majó-Vázquez assistant professor of Communication Science at the VU Amsterdam. The report includes the expert analysis and contribution of the organizing committee members, comprising the KNAW’s grant recipient alongside Mariken van der Velden, professor of political communication of the VU Amsterdam, Felicia Loecherbach, assistant professor at the Amsterdam School of Communication Research and Nicolas Mattis, Postdoctoral Researcher at the University of Amsterdam. 

The symposium had one specific goal: to promote, protect and understand citizens' exposure to diverse news information. And it was organized around two questions. First, what are the effects of the EMFA’s right to media customisation on democratic citizenship? And second, how can we study online news exposure patterns to understand the consequences of that right? 

This report aim to contribute to addressing these questions. They key points covered in it are summarised below: 

• A new regulatory framework pending significant definitions. The EMFA is now actively reshaping the environment in which citizens encounter news. Through its right to media customisation (Article 19) and its safeguards for the visibility of media content on very large online platforms (Article 17), the Act gives citizens and media providers a new lever over exposure diversity. Its practical effect, however, will depend on design and implementation. The new regulatory framework does not cover recommender systems, and it is silent on the minimum options users should have when changing default settings. Without those specifications, the rights’ effects on information exposure is currently undefined. 
• The experts agree that the EMFA’s reach is limited by an unresolved definitional question: “who is the media?”.  Its service-based definition of a media service provider excludes some of the most vulnerable actors in the information ecosystem, and self-declaration leaves the decision to platforms’ discretion. 
• The right of media customization could deepen fragmentation rather than repair it. Because the right could, in principle, narrow rather than widen news diets, its roll-out warrants close monitoring for unintended democratic consequences.
• Access to diverse information doesn't guarantee a more resilient public. A diverse supply is necessary but not sufficient, and its effects on citizens are real but conditional according to the evidence available. Availability of media options does not guarantee engagement, and the same diverse content can inform or mislead, build trust or corrode it, depending on the identities, habits, and peer relationships through which audiences process it. Deliberately engineering desirable outcomes through exposure diversity is genuinely difficult. Experts warn that facilitating one outcome can come at the expense of another, and interventions can even backfire.
• Exposure diversity is a multi-level construct, and how we measure it shapes what we conclude about its effects. It spans topics, actors, and viewpoints, and no single index captures it fully. The choice of measure should follow from the normative goal and the data at hand. Identifying viewpoints, in particular, remains an especially hard methodological challenge for both human and automated approaches.
• Substantial gaps remain in what we know about exposure diversity and its effects. We still do not understand how heavily non-traditional practices including influencers, podcasters, and peer-mediated flows, among others, weigh in citizens’ overall media mix, how they interact with conventional consumption, or how that interaction shapes democratic engagement across social groups. Closing these gaps calls for wider measurement frameworks, attention to how informed citizens feel, multi-method “linkage designs” that also incorporate qualitative approaches