Large language models (LLM) are being used in educational settings to provide guidance, feedback, and policy-related judgments. However, these systems produce inconsistent responses to similar cases, raising concerns about fairness, transparency, and trust. Rather than relying on probabilistic model output, our approach represents cases in terms of salient features and applies principles derived from institutional academic policies to guide decisions. The LLM remains responsible for interpreting cases, while a separate layer using principles ensures similar cases are treated consistently. This approach offers a middle ground between unrestricted generative AI and rule-based systems. By making the basis for decisions explicit and inspectable, it provides greater transparency, supports accountability,and allows policies to be operationalized in a consistent form. We illustrate the approach using examples drawn from our college policies concerning student use of AI and discuss the implications for responsible deployment within the liberal arts.
This hands-on workshop uses a tabletop role-playing game format to build practical AI literacy in an engaging, low-stakes environment. Participants take on characters facing real-world scenarios, such as drafting a research paper, responding to a workplace email, evaluating a suspicious source, and must decide how, when, and whether to use AI tools to help. Dice rolls and simple game mechanics introduce chance and consequence (a "critical fail" might mean an AI hallucination slips into a report; a "success" might mean catching a bias in a generated summary), turning abstract concepts like ethical decision-making, verification, bias, and academic integrity into lived experience rather than lecture material.
Oral assessments are often proposed as an “AI-resistant” alternative to written assignments that foster cognitive skills that are especially important in an AI-infused world, such as empathy, conversational communication, and situated-decision making. The unique opportunities offered by oral assessments, however, also come with unique equity challenges that require creative and potentially unfamiliar solutions. Students’ lack of practice with the format, anxiety and performance stress, language and cultural barriers, and neurodiversity (including differences in processing time, speech, and social fluency) all constitute hurdles for equitable assessment, particularly for instructors who have never designed an oral examination. In this session, we will explore these risks, identify strategies for designing and implementing assignments that reduce their impact, and apply those strategies to participants’ specific instructional contexts. Participants will leave the session with a clearer understanding of the equity challenges associated with oral assessment, knowledge of concrete techniques for addressing them, and at least one new or reformed assignment that incorporates inclusive oral assessment components ready to be implemented into their existing courses.
Colleges and universities have barely begun to evaluate the risks and rewards of ubiquitous AI use on their campuses. This session introduces a framework for assessing AI readiness and risk in divisions, departments, centers, and programs. This presentation will highlight key research, legal precedents, and strategies for identifying gaps between policy and pedagogy. Using an experiential approach, attendees will have space to reflect on the profound AI-related changes taking place and consider which roles or areas are most exposed to AI-driven change. Organized into six readiness domains, the session aims to help participants develop a practical, data-driven approach that supports student learning and academic integrity while managing the changing landscape of faculty and staff work. Participants will also have the opportunity to complete a brief AI readiness and risk survey, which offers a diagnostic starting point for administrative or department heads, program or staff leaders, and task force members.
