Connecticut College AI and the Liberal Arts Symposium

While considerable research has examined how students use generative AI, and how faculty feel about that use, no study to date has reversed the lens. This mixed-methods campus survey of ~400 liberal arts students and 100 faculty investigates how these populations perceive and evaluate each others' AI practices. Our findings reveal striking convergences and divergences. Students and faculty share similar fears (cognitive offloading, accuracy, environmental impact) and similar hopes (more training, experience, and institutional guidance). Yet students express significant dissatisfaction with specific faculty practices. For example, students are particularly against using AI to generate letters of recommendation, significant course content, and grading of written work, despite some faculty reports of using AI to do these tasks. These findings raise urgent questions about the ethical standards we apply to ourselves versus others. We discuss implications for institutional policy, pedagogical transparency, and the equitable development of AI norms in the liberal arts.