In evaluating student work, the risks of getting AI wrong are real, but so are the rewards of getting it right.
In a Chronicle of Higher Education essay, Michelle D. Miller argues that colleges must move quickly to rethink how they assess student learning in the age of generative AI. Rather than waiting for consensus, she outlines four principles for progress: protect faculty innovators, keep working groups narrowly focused, center assessment conversations within disciplines, and let learning goals—not tools—drive instructional choices. Discipline-specific discussions, she notes, help faculty move past abstract fears of AI and toward practical, mission-aligned solutions.
