INFUSE is a full-day workshop on information access in uncertainty scenarios at the European Conference on Information Retrieval (ECIR '26). It aims to bring together the broader information retrieval community and researchers and practitioners from adjacent disciplines to discuss emerging challenges, share diverse perspectives, and explore novel approaches to information access in contexts of uncertainty.

People often encounter or seek information on issues for which there is no clear answer. For example, information may be scarce or nonexistent, as in data voids or unanswerable questions; knowledge might be evolving, as with current events, emerging topics, or rumours; or there may be multiple perspectives on an issue, as with debated topics.

We refer to these as uncertainty scenarios. In such scenarios, information access systems face various risks, including eliciting a false sense of information certainty and over-reliance, promoting misinformation and disinformation, or exposing users to viewpoint-biased information. 

The First Workshop on Information Access in Uncertainty Scenarios aims to provide a forum for the broader IR community to discuss emerging challenges, share diverse perspectives, and explore novel approaches to information access in contexts of uncertainty. Our goals are to:

• Draw Attention to Emerging Risks: To draw attention to emergent issues, such as issues arising from the rapid adoption and integration of LLM-enabled search services, which may lead to potentially harmful consequences.

• Build Community: Bring together researchers and practitioners from diverse backgrounds who are interested in information access in scenarios of uncertainty.

• Exchange and Advance Knowledge: Engage in structured discussions on topics and questions collected by the organisers prior to the workshop, as well as questions that emerge during the workshop.

• Improve Information Access Systems: Contribute to the development of information access systems that prioritize informed citizens and critical thinking, and raise awareness of the complex nuances of information.

• Promote Collaborations: Create spaces for researchers to discover shared interests and explore opportunities for impactful collaboration.

Philipp Lorenz-Spreen
Social Tipping Points: The Spiral of Attention and Platform Power 

This talk moves from everyday attention demands, notifications, news feeds, and social media, to the large platforms that profit from them. Over the past two decades, digital technologies have transformed public discourse: increasing interconnectedness has enabled more self-organized debate, while platforms, their algorithms, and now generative AI have gained unprecedented influence over what we see and share. I argue that power over discourse lies less in direct content control and more in the structural design of platforms. Drawing on complexity and behavioral science, the talk explores why platforms can support democratic outcomes in some contexts while fueling polarization, mistrust, and autocratic resilience in others. Finally, I discuss how we can measure these effects and what alternative platform designs might look like. 

Philipp Lorenz-Spreen is a junior research group leader in the field of Computational Social Science at the Center Synergy of Systems at the TU Dresden. He has also been conducting research at the Max Planck Institute for Human Development since completing his doctorate in theoretical physics at the TU Berlin on the dynamics of collective attention. His research focuses on the complexity of self-organised online discourse, and its impact on democracies worldwide.

Norbert Fuhr
Uncertain, Incomplete and Contradictory Knowledge in IR

Norbert Fuhr is a full Professor at the University of Duisburg-Essen since 2002, leading the Duisburg Information Engineering Group. Before this, he was an Associate Professor at the University of Dortmund, and completed his PhD in Computer Science at the Technical University of Darmstadt. In his research, he focuses on interactive and domain-specific information retrieval, and the validity of information retrieval experiments.