Rapid advancements in artificial intelligence have the potential to transform the US education system, and the field of assessments is no exception. As these technologies continue to evolve, they can be key levers in creating authentic and ecologically valid tools that generate useful data on skills, mastery, and progress for diverse learners who have not been served by traditional assessment. Through unique real-work examples and case studies, this panel will explore the exciting possibilities and the real risks that exist at the intersection of AI and assessment.