Isabelle Hupont – Rethinking QoE in the age of AI: Ethics, human oversight and the role of regulation
Quality of Experience (QoE) in the era of artificial intelligence (AI) demands a broader and more integrated perspective. While QoE research has evolved from purely technical metrics—such as latency or visual fidelity—towards more nuanced, context-aware approaches, a new frontier is emerging: how to account for trust, fairness and human agency in the evaluation of AI-based systems.
Existing QoE methodologies, although increasingly sophisticated, still fall short in capturing the ethical and societal implications introduced by recent advances in AI. The latest generation of AI systems—spanning facial analysis tools, recommender algorithms, and generative models—has rapidly reached mainstream adoption. Unlike earlier AI, these systems are now embedded in everyday applications used by general, non-technical audiences. This shift poses new risks: from algorithmic bias in decision-making to automation bias in human behaviour, impacting users’ perceptions, levels of trust and sense of control.
The European regulatory landscape, particularly the AI Act, provides a relevant framework to understand this shift. It introduces a paradigm focused not on the technology itself, but on the risks associated with its intended use. This purpose-driven, risk-based approach places fundamental rights, user safety and democratic values at the heart of digital policy.
This perspective invites a reflection on how ethical, legal and human-centred dimensions can complement existing QoE approaches, especially in contexts involving vulnerable, diverse or non-expert user groups.
Isabelle Hupont is a Scientific Advisor at the Spanish Ministry of Science and a recognised expert in trustworthy AI and Human-Machine Interaction. With over 18 years of experience across academia, industry, and public institutions—including the European Commission—she combines a strong technological and research background with a solid track record in science-for-policy. She has led and contributed to numerous R&I projects, published extensively and actively promotes ethics and diversity in STEM.
https://www.linkedin.com/in/isabelle-hupont-torres-95747910/
Regina Bernhaupt – From Playful Probes to Interactive Dialogues: Reimagining Research Methods in the Age of LLMs and Digital Twins
Methods like questionnaires, interviews or probing have been used in the area of human-computer interaction for decades. Recent technologies like large language models and digital twins are changing how these methods are used and what we can achieve with them. This keynote explores the evolution of these methods, with a focus on playful probing techniques, examining how they are being redefined by AI’s conversational depth and simulation technology’s ability to mirror reality. Discover the opportunities and challenges that come with these cutting-edge tools and learn how they are shaping the future of inquiry and innovation in research.
Regina Bernhaupt is Full Professor of Measuring and Analyzing Quality of Dynamic Real Life Systems. Making people’s lives better by understanding how, when and why interactive systems fail is the driving force behind Bernhaupt’s research. Her focus is on how to evaluate technologies during all design and development phases – from idea generation to product deployment. In her work, she studies how to evaluate usability and user experience in various contexts, especially for entertainment oriented products and services.
https://www.tue.nl/en/research/researchers/regina-bernhaupt/
Michele Zorzi – QoE/QoS management in communication networks
The concepts of Quality of service (QoS) and Quality of Experience (QoE) are widely used to characterize the performance of communications, networking, and multimedia systems. According to the context, they may be obtained in various ways (analysis, simulation, experiments) based on objective (e.g., throughput or delay) or subjective (e.g., MOS, questionnaires) metrics. Recently, interdisciplinary efforts to characterize the overall performance of these systems, and the need to better understand the relationship between the user experience at the application layer and the network performance at the lower layers, are calling for a clearer definition of the QoS and QoE concepts, and for a study of their dependence on each other. This talk will try to address this issue and to discuss examples of this approach.
Michele Zorzi is a Professor at the University of Padova, an internationally recognized expert and influential researcher in the area of wireless communications and networking, and the PI for many funded research projects both in Europe and in the US. He received many awards and honors, is an IEEE Fellow, and has served three times as Editor-in-Chief of major IEEE periodicals and held several elected or appointed roles within ComSoc (including twice voting member of the BoG).