The Future of Edtech: Building Better AI for Education

Download

Just a few years ago, the conversation about AI in higher education was mostly about detection: catching it, blocking it, keeping it out of student work. Now, more than 90% of students and educators now use AI, and it's moved into the cognitive core of learning, where students summarize, analyze, and reason with it every day.

The strategy for AI in higher ed hasn't followed usage trends. In 2024, only 20% of US institutions had published an AI policy, even as 92% of provosts reported faculty and staff asking for guidance. The result is AI adopted feature by feature, department by department, with no shared standard for what good looks like.

This report gives leaders a way to close that gap. It lays out an institutional audit built around the five risks that surface when AI runs without academic context, the questions to ask in each case, and the characteristics of AI worth adopting. 

What's inside: The audit examines five risk areas:

  • More output, less learning, when students submit polished work without doing the thinking behind it.
  • Poor teaching quality, when generic AI produces content that doesn't match learning outcomes.
  • Equity and accessibility gaps, when AI helps confident users most and widens the distance for multilingual students and those with accessibility needs.
  • Weak data transparency, when no one can see what data an AI system uses or where human oversight sits.
  • Ad hoc adoption, when isolated features pile up without a long-term plan.

For each risk, you get the questions to ask your team and a description of what context-aware AI does differently, including how Canvas by Instructure and IgniteAI approach each one.

How did you hear about us?
Download