The adoption of generative AI has destabilized the traditional artifact economy in higher education, shifting academic integrity discussions into a crisis point.
For decades, higher education has relied heavily on take-home exams, essays, and traditional assessment models as measures of student competence. Now, because modern generative tools can rapidly produce most of that absent of any actual student cognition, the system is breaking. And that forces faculty to rethink how learning is evaluated.
As institutions and individuals grapple with this shift, early adaptations often fall into a risky inequality loop. Well-resourced campuses and small seminars can readily absorb the administrative challenges of high-control, low-AI-use-risk assessments, such as oral defenses, timed in-class writing, and live process evaluations. But large-enrollment gateway courses and commuter campuses frequently lack the staffing and flexibility to make this shift cleanly. And on top of all of that, tenure-stream instructors must navigate these pressures while also balancing traditional performance expectations for promotion.
So what is to be done?
Fortunately, the same technology disrupting traditional assessment can eventually help lower the cost of capturing continuous evidence streams. By treating AI as a tape recorder rather than a ghostwriter, faculty can use AI tools to design low-stakes workflows that circumvent the capacity problem of one faculty member assessing dozens–if not hundreds–of students. Using generative tools may allow professors to automatically track student revision histories. This encourages transparent tool use by students, even in a larger classroom or lecture hall.
Making the shift toward process-oriented evaluation may allow tenure-stream faculty to measure real growth and critical judgment even under constraint.
Chrysanthos Dellarocas, "AI Will Break Assessment Before It Fixes It," Inside Higher Ed, February 19...
This content was created by a human and refined by Gemini.
