AI-Related Risk and Uncertainty
Giacomo Zanotti, Daniele Chiffi, Viola Schiaffonati · Lecture notes in computer science · 2024
Abstract Discussions on the risks involved in the deployment of AI systems are increasingly prominent in both public discourse and scientific debates. While talk of risk plays a crucial role in framing ethical and societal problems related to AI, we argue that it could profitably be associated with a clear analysis of uncertainty. Starting from a multi-component approach to AI-related risk assessment and mitigation, this chapter discusses the way the deployment of AI systems often takes place in contexts in which uncertainty is not meaningfully quantifiable.