Abstract
The term "generative AI" refers to computational techniques that are capableof generating seemingly new, meaningful content such as text, images, or audiofrom training data. The widespread diffusion of this technology with examplessuch as Dall-E 2, GPT-4, and Copilot is currently revolutionizing the way wework and communicate with each other. In this article, we provide aconceptualization of generative AI as an entity in socio-technical systems andprovide examples of models, systems, and applications. Based on that, weintroduce limitations of current generative AI and provide an agenda forBusiness & Information Systems Engineering (BISE) research. Different fromprevious works, we focus on generative AI in the context of informationsystems, and, to this end, we discuss several opportunities and challenges thatare unique to the BISE community and make suggestions for impactful directionsfor BISE research.