What are we telling our students about AI? An exploratory analysis of university instructors’ generative AI syllabi policies

Within higher education, the popularity of generative artificial intelligence (AI) has led many administrators and faculty to stress the need for policies regarding its use. However, given its relative novelty, it is unclear how instructors create and communicate such policies to students, in terms of both content and style. We present a detailed content analysis of university instructors’ generative AI syllabi policies, examining substantive themes, linguistic style, and learner-centered style elements. Using a mixed-methods approach combining human coding and automated linguistic analysis, we analyzed 92 AI use policies from various disciplines and institutions. Results revealed that most instructors crafted policies that defined AI through example and allowed its use with restrictions. Thematic warnings about AI’s limitations were positively associated with perceptions of a warmer, student-centered tone. Linguistically, while the use of second-person pronouns was most common, only first-person and collective pronouns correlated with a warmer tone. Most policies analyzed also incorporated learner-centered elements like rationales and student responsibilities, though fewer invited students to request feedback on the topic of AI. Overall, policies tended to place the primary responsibility on students for ethical AI use. Results are interpreted with a functional approach toward syllabi construction, before offering practical applications for instructors.