AIIRS is a task-classification instrument designed to assess the inherent risk of using generative AI in bounded tasks. It provides a framework based on epistemic dependence, verifiability, and consequences of error to help organizations apply appropriate safeguards within governance and quality-assurance processes.