Generative artificial intelligence (GenAI) is rapidly transforming every stage of the assessment development lifecycle. While traditional psychometrics remain essential for ensuring quality and fairness, GenAI tools increasingly enable faster, more scalable, and more inclusive creation of exams and scales. This session summarizes cutting-edge research and real-world examples of GenAI’s impact on exam development. We present evidence and case studies showing that large language models and other AI methods can: 1) Efficiently generate comprehensive and relevant job tasks, producing lists comparable in quality to those authored by subject matter experts; 2) Draft high-quality, content-valid items for both knowledge and personality assessments, supporting rapid production and diverse item needs; 3) Identify redundant or overlapping (“enemy”) items in large item pools via text embeddings and similarity metrics, streamlining quality control; 4) Facilitate initial review of item bias, clarity, and cognitive complexity, as well as classify items by Bloom’s taxonomy, though human oversight is still vital; 5) Automate the scoring of open-ended responses, providing accurate and explainable ratings that align closely with expert judgments. The presentation will feature succinct overviews of each application, illustrative examples, and practical insights about integrating GenAI into existing assessment workflows. Attendees will gain a clear, evidence-based understanding of how human-AI collaboration can drive more efficient and equitable assessment development, enabling high-quality exams for diverse populations.