THEMIS in ERCIM News: AI for Science Special Edition

Our work has been featured in the AI for Science special edition of ERCIM News, a quarterly magazine published by the European Research Consortium for Informatics and Mathematics (ERCIM). ERCIM News highlights applied research and innovation in computer science, AI, and data science, with a strong focus on work that is relevant to European research, industry, and policy communities.

Cover Page of ERCIM News Special Edition: AI for Science

Our article, “Developing Human-Centred Trustworthy AI as Infrastructure for Reliable Decision Support,” presents insights from the THEMIS 5.0 project on how trustworthiness can be operationalised in real-world AI systems. Rather than treating trust as a single metric, the work argues for a human-centred, risk-based approach that reflects the values, contexts, and needs of different users and sectors.

The contribution discusses practical methods for assessing and communicating trust in AI, including user-aligned trust criteria, narrative explanations, and explicit risk identification. These approaches are particularly relevant as AI systems increasingly function as infrastructure for high-stakes decision support in domains such as healthcare, ports, and media.

You can read the full article in the ERCIM News AI for Science special edition here: https://ercim-news.ercim.eu/en143/special/2998-developing-human-centred-trustworthy-ai-as-infrastructure-for-reliable-decision-support