Adding one "Intro to AI" elective doesn't make a department AI-ready. A genuine framework touches how AI is taught, how it's allowed to be used, and how outcomes are assessed across the whole program.
Map where AI already touches your curriculum
Before adding anything new, audit where AI is already relevant — in research methods, in industry tools students will use after graduation, in assessment design. Most departments find it's more places than expected.
Separate "AI literacy" from "AI specialization"
Every student needs baseline AI literacy — using tools responsibly and effectively. Only some need deep technical specialization. Conflating the two leads to curricula that serve neither well.
Build assessment that survives AI assistance
If an assignment can be fully completed by pasting the prompt into ChatGPT, it was already testing the wrong thing. Assessment design needs to evolve alongside tool capability.
Revisit annually, not once
AI tool capability changes faster than typical curriculum review cycles. A framework that isn't revisited yearly will be outdated within two.