About
Creating Trustworthy AI techniques has become increasingly important. The pressing challenge is to ensure that systems are appropriately trusted by users. This is especially true in high-stakes domains (medical, legal, financial) where both overreliance and underutilization can have significant consequences. As AI rapidly moves from research into real-world applications, it is critical for systems to not only be technically robust, but appropriately trustworthy.
Substantial progress is being made through research, standards, and practice, with diverse perspectives on trust (including normative, functional, and social approaches). Yet, key challenges and questions remain, among which:
- How do we design, evaluate, and verify trustworthy systems?
- How to measure, understand, and calibrate human-AI trust?
- And how can these ideas be translated into practical tools to apply in real-world AI development and deployment?
This interactive workshop addresses these questions through a roundtable discussion format, bringing together participants from diverse backgrounds and areas of expertise. Through structured small-group discussions on pre-defined topics, the session aims to exchange ideas, bridge disciplinary perspectives, and cultivate connections that can spark future collaborations and joint research efforts.
Join us for an interesting discussion and grow your trustworthy network!
Agenda
Intended Audience
This workshop welcomes a broad and interdisciplinary audience, including: students, researchers (of ranging levels of expertise), and industry practitioners.
Participants are invited to contribute their domain-specific expertise, share experiences, discuss challenges, and explore new opportunities for collaboration around Trustworthy AI.
Additional Considerations
- Interactive Format: We emphasize small-group discussions rather than including lectures.
- Output: The workshop will produce a synthesis of findings derived from the discussions, including key insights, identified challenges, and promising directions for future research. This will be compiled into a structured report or white paper, which will be distributed to participants and the BNAIC community.
- Inclusivity: We encourage participation from diverse disciplinary and demographic backgrounds.
Keynote Speaker
Michel Dumontier
Maastricht University
Dr. Michel Dumontier is a Distinguished Professor of Data Science at Maastricht University. His research focuses on the development of computational methods for scalable integration and reproducible analysis of FAIR (Findable, Accessible, Interoperable and Reusable) data across scales - from molecules, tissues, organs, individuals, populations to the environment. His group combines semantic web technologies with effective indexing, machine learning and network analysis for drug discovery and personalized medicine. He coordinates the HE REALM project, which is establishing a regulatory sandbox for AI-based medical software devices. REALM has developed an independent evaluation platform, five demonstrator-led living labs, and an academy that brings together manufacturers, notified bodies, regulators, hospitals, healthcare practitioners, and patients.
Organizers
Daan Di Scala
TNO & Utrecht University
Roos Bakker
TNO & Leiden University
Michael van Bekkum
TNO