The present study investigates how Generative Artificial Intelligence (GenAI) use cases emerge and evolve within organisational settings, challenging dominant "blueprint" perspectives that assume static implementation pathways. While Generative AI offers transformative potential by generating novel content and workflows, research often overlooks the dynamic, socio-technical processes through which its value is realised in practice. Addressing this gap, we conduct a longitudinal case study analysing 18 months of GenAI implementation within the pre-sales processes of a manufacturing firm. Our analysis reveals that GenAI use cases evolve dynamically across two key dimensions: process formalisation (structured vs. emergent) and user type (internal vs. external). Findings indicate use cases emerge organically through user interaction, organizational learning, and strategic imagination, progressing through four distinct phases. This evolution is enabled by specific conditions, including organisational legitimacy, technological flexibility, reflexivity, experimentation, and strategic foresight. The study develops a dynamic framework mapping this use case emergence, identifying key patterns and enabling conditions. Theoretically, this research contributes by shifting focus from static GenAI adoption models to dynamic, path-dependent processes, extending affordance theory through "evolutionary affordances," and deepening understanding of value co-creation and co-destruction dynamics over time. Practically, the framework assists organizations in managing GenAI adoption as an emergent process, guiding capability development and investment strategies for expanding GenAI applications across diverse organizational contexts.