RUME 28
2026
Alexandria, Virginia
Characterizing Generative AI Policies in a University Mathematics Department
Page: 1397
Following the rapid introduction of generative artificial intelligence (GenAI) tools such as ChatGPT, universities have an increased incentive to define expectations for AI use. This study analyzes 14 instructor-written GenAI policies from a university mathematics department. Using inductive thematic analysis, we examine how calculus-sequence and proof-based courses frame GenAI in relation to learning, integrity, and reliability. Most policies took a moderate stance, allowing conceptual use but prohibiting AI-generated solutions to homework, while a few were fully restrictive or open. Instructors more frequently mentioned learning and metacognition than academic integrity or unreliability. Policies did occasionally possess contradictory language, banning GenAI while requiring citation, suggesting ongoing ambiguity. Clarifying such tensions and supporting students in ethical, informed use remain central challenges as disciplinary norms continue to develop.