Seminar Overview
Evaluating the safety and reliability of today’s AI-powered systems has become a crucial aspect to consider before deployment and throughout the entire device lifecycle. In specific regions such as Europe, for high-risk systems, it is currently mandatory to identify and analyze known and reasonably foreseeable risks before the AI system is put on the market, as mandated by the EU AI Act. Although safety, defined as freedom from intolerable risks, is an unmeasurable dependability attribute, assessing it has become a significant challenge in recent times. This involves considering two fundamental factors: the intrinsic uncertainty under golden conditions and the unpredictable behavior caused by random hardware faults affecting the underlying hardware, in other words, their resilience. This talk will address these two themes, providing insights and highlighting representative state-of-the-art works to summarize the safety and reliability challenges of our time and how they are being tackled today.
About Annachiara RUOSPO
Annachiara RUOSPO received the M.Sc. and the Ph.D. degree in computer engineering from the Politecnico di Torino, Italy, in 2018 and 2022, respectively. She is currently an Assistant Professor with the Department of Control and Computer Engineering in the same university. Her main research interests include safety, reliability, and security aspects of AI systems, with a focus on artificial neural networks, and test and verification of modern embedded devices. She is a Member of the AI Existential Safety Community of the Future of Life Institute.