An interdisciplinary contribution at ECAI 2025 explored how different disciplines understand trust in AI and sparked discussions on the social, technical, and institutional dimensions of Responsible AI.
As part of the TRUST-AI workshop at the European Conference on Artificial Intelligence (ECAI) 2025 in Bologna, R-AIMS presented the paper “Rethinking Trust in Responsible AI”. The two-day workshop brought together researchers working on questions of trust, safety, governance, and reliability in AI systems, creating a space for interdisciplinary discussions on trustworthy and responsible AI.
The paper addresses a central challenge in Responsible AI research: although trust is widely recognised as a key principle, its meaning differs considerably across disciplines. While computer science often understands trust in terms of measurable system properties and technical performance, perspectives from sociology, philosophy, and law highlight social, institutional, and normative dimensions. The contribution argues that trust in AI cannot be reduced to a single technical property but emerges through interactions between systems, people, institutions, and societal expectations.
Against this backdrop, the paper proposes understanding trust as a boundary concept – a shared term that supports collaboration across disciplines while acknowledging different perspectives and conceptual approaches. This interdisciplinary perspective resonated strongly with the broader goals of R-AIMS, where technical, social, and ethical questions around Responsible AI are considered together.
Within the workshop programme, the contribution stood out due to its explicitly interdisciplinary and conceptual focus. The presentation sparked interest and discussion among participants, ranging from encouraging reactions and engaged questions to more critical reflections on how trust can be conceptualised and operationalised in AI research.
Beyond the presentation itself, the workshop offered valuable opportunities for exchange with researchers from different institutions and disciplines. In particular a collaborative workshop session on future challenges of trustworthy AI enabled participants to discuss emerging questions in small interdisciplinary groups and exchange perspectives on the development of Responsible AI research.
Alongside the academic programme, ECAI 2025 also provided an opportunity to experience Bologna – a city whose atmosphere and history offered an inspiring setting for scientific exchange.