Communicating Uncertainty to foster Realistic Expectations
via Human-Centered Design
The increased capabilities of black-box AI models and, in particular, large language models (LLMs), have brought forth the need for trustworthiness and explainability for their adoption in real-world tasks. Communicating system uncertainty is essential for achieving transparency and can help users calibrate their trust in, reliance on, and expectations from an AI system. This issue is exacerbated with LLMs which can generate convincing content, leading users to overestimate their certainty. However, uncertainty communication is plagued by challenges such as cognitive biases, numeracy skills, calibrating risk perception, and increased cognitive load, with research finding that lay users can struggle to interpret probabilities and uncertainty visualizations.
The discrepancy between user expectations and system capabilities can result in poor user experience, lower usage willingness, and poor task performance. From a visualisation-driven uncertainty representation on mobile screens through quantile dotplots, to a text-based frequency or first person uncertainty communication via LLMs to an embodied uncertainty communication through an agent's hesitation gestures, interfaces play a central role in enabling personalized, task-specific and interactive uncertainty communication across devices and modalities. This challenge is even more pronounced in agentic systems, where users need to navigate uncertainties across multiple agents without direct access to their actions and interactions.
CURE 2026 workshop will investigate how intelligent user interfaces can aid in bridging the gap between system uncertainty and users' expectations. The workshop will discuss and disseminate novel research on designing, developing and evaluating intelligent user interfaces for uncertainty communication. CURE 2026 aims to gather researchers from diverse fields, such as interface and interaction design, Human-Computer Interaction (HCI), personalization, data visualization, and cognitive science, as well as domain experts from fields where uncertainty plays a vital role such as healthcare, intent modelling and financial forecasting.
Important Dates
| October 10, 2025 | Submission Window Opens |
| December 19, 2025 | Paper Submission Deadline |
| February 02, 2026 | Acceptance Notification |
| March 23, 2026 | (Anticipated) Workshop Date |
Scope and Topics
We invite submissions on the following topics:
- Studies on the effects of uncertainty communication on user trust, engagement and interaction
- Novel interface design techniques for communicating uncertainty
- Challenges, pitfalls and opportunities of successful uncertainty communication
- Theoretical work on taxonomies for uncertainty communication
- Human-in-the-loop uncertainty and expectation calibration
- Benchmarks and metrics for robust evaluation of uncertainty communication
- Uncertainty communication in diverse mediums (e.g. visual, textual, multi-modal)
- Communicating uncertainty and expectation calibration in human-LLM interactions
- Case-studies of uncertainty communication in real-world environments
Submission Information
Submissions can be either short or provocation papers:
- Short papers: Up to 6 pages (excluding references): intended for presenting mature or completed research
- Provocation papers: Up to 2 pages (excluding references): intended as position papers or early findings that aim to offer an innovative perspective on uncertainty communication.
All submissions should follow the CEUR-WS submission format . Accepted short papers will be part of the CEUR-WS proceedings.
Submissions must be anonymized for double-blind review. Reviewing will follow the standards of IUI, with evaluation based on novelty, technical depth, clarity, reproducibility, and potential impact. Accepted papers will be presented as either posters or talks. At least one author of each accepted paper must register and attend the workshop.
Organizers
Program Committee
We are grateful to the following people for helping make the CURE workshop a success:
- Abhraneel Sarma, Northwestern University, USA
- Alex C. Williams, AWS, USA
- Alok Debnath, Trinity College Dublin, Ireland
- Danielle Albers Szafir, University of North Carolina - Chappel Hill, USA
- Eoin Delaney, Trinity College Dublin, Ireland
- Eelco Herder, Utrecht University, Netherlands
- Francesca Naretto, University of Pisa, Italy
- Hariharan Subramonyam, Stanford University, USA
- Hyo Jin Do, IBM Research, USA
- Ivana Dusparic, Trinity College Dublin, Ireland
- Jacy Reese Anthis, University of Chicago, USA
- Jan Leusmann, LMU Munich, Germany
- Jeroen Ooge, Utrecht University, Netherlands
- Justin Edwards, University of Oulu, Finland
- Leon Fröhling, GESIS Leibniz Institute for Social Sciences, Germany
- Maxwell Szymanski, KU Leuven, Belgium
- Michael Hedderich, LMU Munich, Germany
- Owen Conlan, Trinity College Dublin, Ireland
- Prakash Jamakatel, University of the Bundeswehr Munich, Germany
- Rune Møberg Jacobsen, Aalborg University, Denmark
- Samuel Rhys Cox, Aalborg University, Denmark
- Selene Baez Santamaria, University of Zurich, Switzerland
- Siddharth Swaroop, Harvard University, USA
- Simret Araya Gebreegziabher, University of Notre Dame, USA
- Siya Kunde, IBM Research, USA
- Srishti Yadav, University of Copenhagen, Denmark
- Sunnie S.Y. Kim, Apple, USA
- Timo Speith, University of Bayreuth, Germany
- Yan Zhang, University of Melbourne, Australia
- Yue Fu, University of Washington, USA