The University of Sydney has created a menu of assignment options for instructors that effectively allows students to use generative AI on most out‑of‑class work while protecting supervised assessment. The redesign is an effort to restore assessment validity and provide clearer faculty controls over where and how AI is permitted. The menu approach gives instructors a structured way to differentiate in‑person, invigilated tasks from take‑home assignments where AI can be leveraged. University officials say the change acknowledges pervasive student AI use and shifts focus to assessment design rather than blanket bans. For teaching and learning leaders, the move signals a pragmatic model for reconciling academic integrity with technology adoption—one that other institutions may adopt to manage risk while preserving learning outcomes.