Hybrid decision systems combining human expertise and AI are increasingly prevalent. While they offer benefits like improved efficiency and quality, challenges concerning responsibility, bias, and transparency also arise. THEMIS 5.0 brings together experts from multiple disciplines to ensure these systems are trustworthy, ethical, and meet specific domain requirements.
The project will create a framework for enhancing the trustworthiness of AI decision support systems. It includes assessing fairness, robustness, and decision impacts, involving domain experts in the assessment process. SINTEF's role involves developing processes and tools to balance human and AI involvement in decision-making, tailored to each situation's demands and risks.
SINTEF will also work on enhancing AI decision support accuracy and robustness, making it simpler for experts to evaluate the contributions of AI. This includes implementing AI-driven anomaly detection and activation analysis for clearer explanations and decision-making adjustments.
THEMIS 5.0 will explore its approach in three sectors: healthcare, logistics, and media, focusing on fairness, robustness, and decision impact assessment with expert involvement. The project involves 16 partners from 10 countries, led by Gruppo Maggioli and funded by the Horizon Europe program for research and innovation.
Funding
This project has received funding from the European Union’s Horizon Europe research and innovation programme under grant agreement No 101121042.