The AI, Media and Democracy ELSA Lab researches the impact of AI on the democratic functioning of the media. Together with journalists, media professionals, designers, citizens, researchers and public and societal partners, the Lab develops and tests value-driven, human-centred AI-applications and ethical and legal frameworks for responsible use of AI. The aim of the Lab is to stimulate innovation of AI applications that strengthen the democratic function of media.

The goal of this ELSA lab network project is to contribute to the realisation of impact with ELSA labs. The network project will facilitate ELSA lab knowledge production and dissemination and investigate the lessons learned from the construction and first developments of the ELSA labs. The lessons learned will be translated into an ELSA lab blueprint consisting of a set of guiding principles regarding ELSA lab collaboration, methods, impact and sustainability. Knowledge production and dissemination will be facilitated by an ELSA lab learning network, for which this network project will lay the foundation.

Throughout the last decade, concerns have been growing over the (lack of) transparency of digital infrastructures. Notably, (academic) researchers have decried the many obstacles preventing them from observing and scrutinising privately-held data. At the same time, the EU regulator has put forward a barrage of legislation purported to boost transparency and facilitate data access, primarily in light of internal market goals. Yet, data access for research only features in the margins of these (proposed) data regulations, if at all. Still, the observability of digital infrastructures is vital, not just from a market perspective or for holding power to account, but also for scrutinising the (individual and societal) impact of these infrastructures and observing the world in general. There is a clear need for critical and global reflection on how new data access rules shape research agendas but also on the positionality of different actors in these debates. This book provides a platform for critical reflection on these matters. Embracing the complexity and multifacetedness of the questions at hand, we brought together a group of scholars from the following communities and fields: critical legal studies, science and technology studies, law and political economy, critical data studies and digital humanities. The book aims to bring together critical perspectives on academia’s data production, data use and data ownership, and also explore what kinds of responsibilities exist for academia in data access regimes, beyond legal and methodological duties. In doing so, the book will examine the implications of regulators outsourcing accountability mechanisms to academia via access regimes, the agenda-setting power afforded by data access regimes and the type of dependencies created by data access on academics/ researcher communities.

Our project explored how to make news recommendations more diverse and responsible. We developed new ways to define and measure diversity in news using cutting-edge language models, tested how small design changes (or “nudges”) can guide users toward broader content without backlash, and created an annotated dataset of Dutch news viewpoints. Our results informed Dutch media organizations and EU policy and produced two PhD dissertations, several studies, and open-source tools and data.

Smartphone users in Europe and the United States use apps en masse. The underlying ecosystem, Apple iOS or Android, determines how these apps work and plays an important role when it comes to providing transparency about what happens to users data and privacy. This project aims to determine how, in a technical sense, transparency is influenced by the ecosystem and how the different legal frameworks play a role in this. The findings of the study will be based in part on an intensive user study.