Michela Ghezzi

I completed my undergraduate studies between Italy and Finland, with academic training in Sociology at the University of Milano-Bicocca and further coursework at Åbo Akademi University in Turku. This integrated programme combined sociological foundations with training in social psychology, social robotics, and human-robot interaction, providing a coherent perspective on the social and cognitive dimensions of human-technology relations.

After graduating with first-class honours, I completed a first postgraduate degree, a MSc in Cyberpsychology, at Nottingham Trent University in the United Kingdom, where my training focused on the psychological dimensions of digital and immersive technologies, with particular attention to socially assistive robotics and user experience in human-robot interaction. I subsequently completed a second postgraduate degree, an MA in Philosophy of Artificial Intelligence at Northeastern University, which provided a structured engagement with the epistemological, ethical, and conceptual foundations of artificial intelligence.

I am currently a PhD candidate at Eindhoven University of Technology, where I conduct research at the intersection of artificial intelligence, philosophy, and cognitive science within a multidisciplinary academic environment.

My research addresses foundational questions concerning what it means to evaluate an AI system in epistemically informative ways. Rather than focusing exclusively on observable performance, I investigate how different levels of analysis -behavioural, capability-based, and mechanistic- contribute distinct kinds of evaluative insight.

In particular, I examine the role of internal structures and processes in grounding robustness, reliability, and the possibility of principled intervention, arguing that evaluation practices remain incomplete when they rely solely on external or outcome-based criteria. Drawing on philosophy of science and cognitive science, my work aims to articulate evaluation frameworks that make explicit the assumptions, explanatory commitments, and epistemic standards underlying contemporary AI assessment practices.

My work is conducted in close collaboration with the Philosophy and Ethics (P&E) group at TU/e, and closely interfaces with ongoing technological developments within the Mobile Perception Systems Lab (MPS Lab). In addition, I am an active member of the Eindhoven Artificial Intelligence Systems Institute (EAISI), as well as the Long-Term Programme ROBUST consortium, a national research initiative funded by the Dutch Research Council (NWO) and NXP Semiconductors.