February 19, 2026
Inside Higher Ed recently published an article by Chrysanthos Dellarocas, Shipley Professor in Information Systems, arguing that generative AI is breaking traditional assessment by making it harder to tell whether student work reflects real understanding.
Essays and take-home assignments no longer reliably measure learning, forcing institutions to rethink how they evaluate students. Dellarocas points to a shift toward tracking the learning process – how students think, revise, and apply feedback – rather than relying on final outputs. While this could improve assessment long term, he warns it may also widen inequality between institutions with differing resources.
Ultimately, AI isn’t just disrupting education, it’s pushing it toward a more accurate way of measuring what students can actually do.
