SIDLIT 2026

As generative AI becomes easier for students and educators to access, many institutions are responding with policies, permissions, restrictions, and detection tools. While these conversations matter, they often miss a central instructional design question: What learning is the assignment supposed to protect?
This session reframes AI guardrails as a learning design issue rather than a compliance checklist. Participants will examine how AI can support learning without replacing the thinking, decision-making, practice, and evidence students are meant to develop. Using practical examples from classroom and online learning contexts, the session will introduce a guardrails audit that helps educators identify which parts of a task may be AI-supported, which parts must remain student-owned, and what evidence can make student thinking visible.
Attendees will consider how guardrails can support academic integrity, accessibility, student agency, and meaningful engagement without relying only on surveillance or tool bans. The session is designed for educators, instructional designers, faculty/staff support professionals, and technology leaders who are helping others make responsible decisions about AI use in learning environments.
Participants will leave with adaptable questions they can use to review assignments, discussions, projects, and assessments at their own institutions or organizations.