Fluorescent materials have been recognized as a series of important luminescent materials, which has important application in display, bioimaging, and chemical sensing. Conventional synthesis of fluorescent materials often relied on the trial-and-error method, suppressing the development of this realm and prolonging the development cycle. Generally, the recent advances in artificial intelligence (AI) technology have brought new opportunities to chemistry, furnishing the transformation of fluorescent molecular design from experience-based to data-driven approach. In this context, several key questions arise regarding AI‑empowered design and development of novel fluorescent dyes. What generative models and training strategies are applicable? What high‑quality datasets are commonly used? How should molecular descriptors be selected? What are representative examples of inverse‑design strategies? To address these questions, this article first outlines the fundamental principles of generative AI and inverse design. Starting from the key step of complex material development-molecular design, we summarize several commonly used generative models for molecular design and compare their strengths and limitations. Subsequently, we introduce the prediction of key properties of fluorescent dyes based on commonly employed fluorescent dye databases and molecular descriptors. Finally, we review domestic and international progress in applying generative AI to fluorescent dye design and discuss potential future directions. By focusing on actionable research pathways, we aim to provide a reference for establishing AI as a reliable engine driving the design and discovery of fluorescent dyes and functional molecules.