Faculty and assessment leaders are reframing AI use from a purely cheating-prevention problem to a learning-outcomes and competency question, as generative tools can produce artifacts without proving understanding. One report describes a faculty moment after a student used ChatGPT to generate a data model that resembled course diagrams, prompting concern that students and peers could not verify whether the output was correct. The broader argument: institutions need to decide what students should be able to do in an AI-saturated environment and whether course designs still build those capabilities. The emphasis shifts toward redesigning learning outcomes and instructional tasks so the mental models behind performance are demonstrable, not merely outsourced to AI. Separately, guidance on “responsible AI” in assessment stresses balancing AI innovation with academic integrity, noting that many students use AI infrequently and often for low-level tasks rather than concept mastery.