AI-powered perception is reshaping vehicle safety, pushing the limits of established functional safety standards like ISO 26262. This panel explores the fundamental tension between AI’s probabilistic, “black-box” nature and the deterministic assumptions behind current FuSa frameworks. Key stakeholders across the supply chain will discuss how emerging guidelines, such as PAS 8800, seek to bridge this gap by evolving testing methodologies and safety assurance processes. Topics include validating opaque AI models, integrating simulation within FuSa workflows, and balancing innovation with regulatory compliance. Attendees will gain insights into practical strategies for building robust safety cases amid uncertainty, advancing automotive safety standards, and ensuring trustworthy perception-driven decision-making in increasingly complex systems.