This chapter investigates and reconsiders the role of Artificial Intelligence (AI) in Interactive Digital Narrative. In the first two sections, a theoretical view of AI in Interactive Narrative will be outlined in terms of of the dichotomy of actions selection or knowledge representation before placing a focus on the difficulties of producing convincing narrations. Two research tendencies will be mapped – one focusing on in-depth investigation of limited components, the other aiming at building an entire user experience. In the third section, user interfaces are integrated in the theoretical approach – as they are often considered late in the design process, as an afterthought on top of an existing narrative engine. While this approach has proven to be efficient for progress in AI algorithms for interactive narrative, it no longer holds if a complete and convincing user experience is the goal. The fourth part will focus on authoring, showing how it is distributed among several design roles: i.e. the creative author, the user-interface designer and the system designer - whose role cannot be restricted to a «tool builder» and whose importance has been understated so far. The final section will loop back to the AI algorithms themselves. It will argue for a re-consideration of the role of AI in interactive narrative. Rather than considering AI as a means to completely automate narrative processes - a strategy that has not lived up to its promises - the role of AI is described as creating generative computing elements that constitute creative material that both scientists and artists need to work on.