The Breakdown
Having a policy isn’t the milestone. The data makes that clear. What separates institutions getting results from those that are stuck is what happens after the policy is written.
A key finding: faculty who use AI daily are nearly four times more likely to report a decrease in workload compared to monthly users. The workload benefit doesn’t spread evenly. It concentrates at the threshold of daily practice. Institutions that haven’t moved faculty past occasional experimentation are absorbing the disruption of AI without yet seeing the relief.
There’s also a shadow AI problem the data surfaces. Paid AI usage has increased year over year across all groups, from 19% to 34% among administrators, 15% to 24% among faculty and 29% to 39% among students. Faculty and students paying out of pocket for consumer tools are operating outside institutional systems, creating data privacy risk and limiting collaborative use.
The report segments institutions into three groups: First-movers (32%), who have a rolled-out institution-wide policy; Builders (55%), actively working on one; and Observers (9%), who don’t expect to develop one. The gap between these groups isn’t just in policy. It’s in philosophy. First-movers treat AI as a literacy imperative. Observers are still managing it as a risk to contain.
Dr. Ford sees the administrator adoption shift as deliberate. “I remember a year ago talking to administrators of campuses who said, ‘We want to be an AI-first campus.’ And my call to action to them was, ‘Well, then you have to be fluent in the tools you’re asking your faculty and students to use.’ I think that shift is probably purposeful because there’s an appetite where before it was, ‘AI is here, it’s tentative, let’s wait and see.’ Now we’re in the era where AI is here to stay.”
She’s equally direct about what separates institutions seeing results. “Faculty who are more likely to be innovative and include AI in their evaluation and assessment, there’s a correlation between that and positive, pro-AI governance on their campuses. So, if you haven’t made a stance, or you’ve created policies that are prohibitive or created barriers around AI, and at the same time say you want to be an AI-first campus, there’s no correlation there.”
Dr. Zone emphasizes that the policy challenge here is unlike any other. “Because AI is changing so quickly, we cannot use the same frameworks for policy building that we have with other historically well-known policy frameworks. But if you have a campus where there are a lot of Status Quo or Defender faculty, it’s going to be harder to have those policy conversations at the implementation level, because people won’t really know how to give feedback if they haven’t experienced it.”