AI is already influencing how assessment organizations think about exam development and psychometrics—but how far should the field go? This debate will explore the opportunities and risks of using AI for job task analysis, item and form development, item banking, standard setting, DIF analysis, equating, and technical reporting. Panelists will examine whether AI can responsibly improve efficiency, consistency, scalability, and item bank health, while also addressing concerns related to validity, fairness, transparency, security, bias, documentation, and human oversight. Attendees will hear competing perspectives on where AI may add value, where it may create unacceptable risk, and what governance practices are needed before AI can be trusted in high-stakes assessment programs.