The role of news recommendation systems in digital democracies

Personalised and algorithmic news recommender systems are increasingly influencing citizens’ media consumption. In the long term, they alter public opinion-forming processes and the structure of journalism. The research team investigated how such recommender systems affect journalistic news production, public perceptions of and trust in journalism. The team also looked at the role of a responsible design of such recommender systems on both the production and the user side.

The increasing automation and personalisation of communication processes in digitalised societies have triggered a public debate on the democratic implications of algorithmic news or recommender systems such as social media, search engines or news aggregation services. However, very little was known about how such systems are implemented in websites or apps by news media themselves and how this impacts the work of media professionals, the strategic processes of media organisations and the expectations and preferences of citizens.

The main goal of this project was to understand how algorithmic news recommender systems influence the work of media professionals, public perception and trust in journalism. Furthermore, we aimed to develop evidence-based recommendations of public interest for improving the application and implementation of news recommender systems.

The project underscores the cautious approach to the adoption of news recommender systems (NRS) by media organisations, balancing technological innvoation and personalisation with editorial criteria and control. It highlights the uncompromised importance of journalistic standards, transparency, and user control to foster internal support for new technologies and maintain public trust in journalism. Overall, the findings point to a slow and responsible implementation of NRS into news media, as both consumers and producers of news place higher importance on traditional, editorial news values and curation. The project also advocates for comprehensive education on algorithms and AI for both media professionals and the wider public in order to address knowledge gaps and misconceptions, ensuring that digital transformation benefits society while safeguarding public values.

Three main messages

1. Although many news media organisations have implemented recommender systems on their websites, their use remains experimental and is marked by various challenges and trade-offs. These include balancing technological advancements with editorial integrity, meeting user expectations while upholding journalistic responsibility, and managing the growing influence of tech players in newsrooms.
2. To foster greater acceptance of news recommender systems, it is essential to align them with journalistic criteria, clearly communicate user benefits, and address concerns regarding filter bubbles and data misuse.
3. Looking forward, users remain sceptical about algorithmic and AI-driven technologies for news-related purposes. Their attitudes toward generative AI mirror those toward news recommenders, underscoring these findings’ relevance for the journalistic adoption of related technologies in the future.

The Role of News Recommender Systems in High-Choice Information Environments

• Prof. Frank Esser, Institut für Kommunikationswissenschaft und Medienforschung (IKMZ), Universität Zürich
• Prof. Sina Blassnig, Université de Fribourg, Fachhochschule Graubünden (FHGR)
• Prof. Aniko Hannak, UZH Digital Society Initiative (DSI), Universität Zürich
• Prof. Claes H. de Vreese, Amsterdam School of Communications Research (ASCOR), University of Amsterdam