This chapter addresses the unique challenges of eliciting ethical and legal requirements for artificial intelligence (AI) systems, emphasizing their dynamic, context-sensitive, and interdisciplinary nature. Unlike traditional software engineering, AI development operates within ethically sensitive and evolving environments, where principles such as fairness, transparency, and accountability are abstract, contested, and difficult to operationalize. To bridge this gap, the chapter proposes Goal-Oriented Requirements Engineering (GORE) approaches, specifically using KAOS models to translate high-level ethical objectives into measurable system requirements. It introduces enriched user stories that explicitly integrate ethical and legal goals, providing a practical template for embedding normative considerations into functional design.