Generative artificial intelligence (GAI), powered by large language models (LLMs), is rapidly transforming fields such as business, healthcare, manufacturing, and education. In educational contexts, these systems can generate instructional content, provide on-demand feedback, enhance student engagement, and personalize learning experiences. Within educational measurement, researchers have begun to leverage GAI to perform tasks that traditionally require intensive human effort. Current studies have explored several distinct roles of GAI in measurement research and practice, including GAI as respondent, test or item developer, rater, and interactive agent. This chapter synthesizes recent advances across these roles, highlights key empirical findings, examines existing limitations, and outlines directions for future research on the responsible integration of GAI into educational assessment.