Assessment is where AI’s impact on teaching is most visible right now. Faculty have split into three distinct groups in response: Integrators (24%), who redesign assessments around AI and report better student engagement; Defenders (23%), who revert to proctored formats and see no equivalent gains; and Status Quo (54%), who haven’t significantly changed their approach. Writing and editing assignments is the number one student use case for AI.
0%
faculty
0%
students
The Integrator/Defender split isn’t just about assessment strategy. It reflects a fundamentally different response to the same pressure. Integrators use AI daily themselves. Defenders mostly don’t. That experience gap shapes everything else.
Dr. Ford is direct about what the data is really saying. “Every time we talk about cheating, I cringe a bit. The conversation around cheating is as old as teaching and learning has been part of our institutional systems. The difference here is that we have to fundamentally shift how we think about evaluation and assessment. We can no longer just focus on these outputs and these products.”
She connects the divide to a deeper pedagogical question. “The conversation should be about how do we engage learners in ways that we can think about evaluation very differently, where co-construction of learning can happen, where the process matters as much as
the product.”
Dr. Zone brings a psychometrics lens that reframes the stakes entirely. “What AI has really called into question is whether what we’ve been measuring ever actually measured what we intended to measure. That is a deeply uncomfortable question for a lot of faculty. But it’s also an opportunity to design assessments that get closer to what we care about, which is whether students can think critically, apply knowledge, and work through complex problems.”
She also notes that faculty who engage deeply with AI themselves are the ones driving assessment change. As Dr. Zone puts it, the experience of using AI changes how you think about what’s worth assessing, and faculty who attend AI course redesign training deploy roughly twice as many assessment approaches as untrained peers.
The institutions getting better outcomes from AI in assessment aren’t the ones with the strictest policies. They’re the ones helping faculty move from Defenders to Integrators. Institutions can:
Invest in AI course redesign training specifically. Faculty who attend this training modify twice as many assessment practices and show higher confidence in AI’s instructional value.
Build toward the Integrator playbook: project-based assessments, iterative assignments with feedback loops, tasks that ask students to engage critically with AI outputs.
Address the design and grading time barrier directly. Unintended AI reliance, design time and grading time are the top challenges both Integrators and Defenders report. Platform support and instructional design resources that reduce these barriers are the practical pathway from Defender to Integrator.
D2L Lumi Remix lets faculty revise existing content with AI, adjusting reading level and structure in real time, cutting the design and grading work that redesigned assessments usually take.