Every assignment can sit somewhere on a spectrum. At one end are AI-resistant tasks, designed so a tool does not help much, often in-class, oral, or tightly local. At the other end are AI-integrated tasks, where students use AI openly as part of the work and are assessed on how well they direct, evaluate, and build on it. In the middle are tasks where limited, disclosed use is fine.

The point is to choose on purpose for each assignment, based on what you are trying to teach, rather than landing somewhere by accident. Then tell students clearly where this assignment sits and why. Lesson 5 develops the integrated end, and Lesson 6 turns these choices into a course policy.