Workshop on Human Centric AI

The ENFIELD Workshop on Human-Centric AI took place on 9 September 2025, at the National University of Science and Technology POLITEHNICA Bucharest, Central Library.

About the Workshop

As AI systems are increasingly used to support decision-making in complex and dynamic environments, ensuring that these systems are understandable, trustworthy, and aligned with human values remains a critical concern.

Organized by the ENFIELD project: European Lighthouse to Manifest Trustworthy and Green AI, the workshop focused on exploring the integration of AI-based decision systems into organizations, with a strong emphasis on transparency, accountability, and human oversight.

The event showcased work carried out within the ENFIELD research project while also welcoming contributions from other research projects, initiatives, and interdisciplinary efforts addressing challenges in human-centered artificial intelligence. The workshop brought together researchers, practitioners, and stakeholders to share insights, methodologies, and case studies that advance human-AI collaboration, explainability, and trust.

Key areas of focus included:

  • Novel explainable AI methods for decision-making
  • Methodologies for evaluating human-AI shared decision-making
  • Interpretable data-driven decision support systems
  • Societal impacts of AI-aided decision-making

In addition to advancing research and interdisciplinary solutions, the workshop facilitated exchanges and collaboration while reflecting on the implications and future directions of research on human-centered AI.

Program

9:30-9:50Reception of participants
9:50-10:05Adina Magda FloreaPOLITEHNICA Bucharest Welcome message
10:05-10:20Cecile HuetEUROPEAN AI OFFICE – DG CNECT Head of Unit “Excellence in AI & Robotics” Opening remarks
10:20-10:50Gregor BaerEindhoven University of Technology Evaluating Evaluation Methods: Class-Dependent Effects in XAI Metrics
 https://arxiv.org/abs/2506.11790
[Abstract]
10:50-11:20Sebastian Heil, Xavier Carpent, Steven FurnellChemnitz University of Technology, The University of Nottingham A Vignette-based Investigation of User Trust in AI
[Abstract]
11:20-11:40Coffee break
11:40-12:10Jules LeguyInstitut Mines-TélécomA Protocol to Study the Impact of XAI on the Performance of AI-assisted Decision Making
[Abstract]
12:10-12:40Alexandru SoriciPOLITEHNICA Bucharest Experience-based Agent Communities in Support of Smart Environment Hypermedia MAS
[Abstract]
12:40-13:40Lunch
13:40-14:10Vassilis VoulgarakisOkys LTD, HumAIne Project Human-in-the-Loop AI for Trustworthy Decision Support: Insights from the HumAIne Pilots
[Abstract]
14:10-14:40Mihai Nan, Cătălin Mihail ChiruPOLITEHNICA BucharestExplaining Human Action Recognition
[Abstract]
14:40-15:10Darian Onchiș, Codruța IstinWest University of Timișoara From Explanation to Unsupervised Segmentation: Fusion of Multiple Explanation Maps for Vision Transformers
[Abstract]
15:10-15:30Coffee break
15:30-16:15Astik Samal, Alberto CasanovaMaggioli S.p.A. Policy Recommendations for Human-Centric AI
[Abstract]
16:15-17:00All participantsOpen Debate The Future of Human-Centric AI: ENFIELD Perspectives

Gallery