The Verb Is No Longer Enough: A Framework for Explicit Learning Expectations in an AI-Assumed Environment

Course-level learning outcomes in post-secondary education typically leave the expected level of human performance implicit. Historically, constrained assessment environments enforced those expectations without requiring them to be named. Generative AI has removed those constraints, exposing a structural gap in how outcomes are written and assessed. This paper presents the AI-Assumed Learning Framework, a five-domain model developed at Holland College that responds to this gap by making learning expectations explicit at the outcome level. The framework distinguishes between expectations requiring automaticity, tool-supported application, authentic expression, critical judgment, and AI collaboration. Application of the framework to learning outcomes from two post-secondary courses demonstrates that the gap is consistent but takes different forms depending on course type. The paper argues that outcome-level explicitness is a more sustainable response to AI integration than assessment-level restriction, and offers practical guidance for curriculum designers and instructors redesigning CLOs in an AI-assumed environment.