Session
Explainability and Human Oversight in AI Systems
Wednesday July 8, 2026 | 10:30-11:30 CEST
Albert II Auditorium
Description
This session brings together researchers and practitioners to examine the state of explainability and human oversight in AI systems, with a focus on both theoretical challenges and real-world deployment. Speakers from academia and industry will discuss the design and evaluation of explainability, addressing how transparency requirements shape AI development in practice. Applied use-cases from industrial settings will ground the discussion in concrete scenarios where explainability directly affects trust, accountability, and decision-making. The session closes with an open debate on the challenges of explainability and human-in-the-loop design, inviting the audience to engage with perspectives from multiple institutional and disciplinary backgrounds.
Speakers
Johan Loeckx
Assistant Professor, Vrije Universiteit Brussel and Head of Research CollectiveUp
Johan Loeckx is a highly experienced AI strategist and technical expert dedicated to ensuring successful, purpose-driven AI implementation. With an AI journey beginning in 1992, decades before the recent surge, Johan possesses a depth of knowledge that transcends current trends, focusing on combining deep technical understanding with strategic vision.
Holding a MSc in Engineering, a MSc in AI, and a PhD from K.U.Leuven, Johan’s foundational expertise is indisputable. This technical command is demonstrated by his early career achievements, including co-designing the patented Belgian encryption system for health information exchange. His capability to both code and think strategically was refined over ten years managing the applied R&D team at the Artificial Intelligence Lab at the Vrije Universiteit Brussel, where he navigated more than 35 real-life projects with external partners.
Géza Németh
Professor and Head of Laboratory, Budapest University of Technology and Economics, Hungary
Géza Németh is a pioneering professor at the Budapest University of Technology and Economics (BME), an “engineer of interactions.” He is a leading figure in the development of numerous innovative interactive solutions. He has been the Hungarian National Contact Point (NCP) for the Horizon Europe AI Network of Excellence, the ENFIELD (www.enfield-project.eu) project, and the head of the international relations working group for the Hungarian Artificial Intelligence Coalition (https://miagyakorlatban.hu/). His research focuses on speech information systems, multimodal, mobile and robotic human-computer interfaces within the field of artificial intelligence, ranging from basic research through development to applications.
Isel Grau
Assistant Professor, TU/e
Isel Grau is a tenured Assistant Professor at Eindhoven University of Technology (TU/e). Her research focuses on explainable AI, interpretable recurrent neural networks, and time series analysis, with a strong interest in building AI systems that are both transparent and human-centric. She co-leads the Human-Centric AI pillar within the ENFIELD project and has contributed over 85 peer-reviewed publications to the field.
Mohammed Al-Radhi
Researcher, Budapest University of Technology and Economics, Hungary
Dr. Mohammed Salah Al-Radhi is a Researcher at the Budapest University of Technology and Economics (BME). His work focuses on developing AI-based methods for speech and neural signal processing, with a particular interest in future brain-computer interface systems. He is also involved in research and supervision activities connecting AI, healthcare, and human-centered communication technologies. In ENFIELD, he contributes to activities related to trustworthy and human-centric AI.
Jeriek van de Abeele
Research Scientist in Analytics and AI, Telenor
Jeriek Van den Abeele joined Telenor Research & Innovation as a Research Scientist in 2020, following a PhD in theoretical particle physics at the University of Oslo. During his PhD, he worked on topics including Gaussian process regression for particle cross-section calculations and genetic algorithms for computational drug development. At Telenor, he has been involved in work on using reinforcement learning for network automation and technical advisory regarding AI governance and the EU AI Act. He is currently leading Telenor’s efforts in the Horizon Europe project ENFIELD, focusing on Green, Human-centric, and Trustworthy AI research directions, including work on building defences against LLM jailbreak attacks. [Photo credit: Erik N.H. Krafft (Krafftwork Photo and film / Flow Event)]