Hao Song, ASPPB
Ting Wang, American Board of Family Medicine
Explore how generative AI and related tools can support every stage of blueprint development—from updating competency frameworks to implementing new exam designs. This session features real-world applications from three medical licensure programs that used AI for literature synthesis, item-to-blueprint alignment, and AI-assisted item generation. Learn how natural language processing and large language models were integrated into workflows to enhance efficiency and scalability, while preserving expert oversight. Presenters will share practical strategies, implementation insights, and challenges encountered, including prompt design, classification accuracy, and quality control. Whether you're new to test development or experienced in blueprint design, you'll gain a step-by-step understanding of how AI can—and cannot—be used responsibly in high-stakes exam programs.