Most colleges have responded to AI the way early factories responded to electricity: by plugging in the new technology without touching anything else.
The result? Modest gains at best, mounting confusion at worst.
Students navigate a patchwork of conflicting policies across departments, faculty worry about plagiarism and dependency, and the fundamental architecture of a college education—how learning is structured, measured, and delivered—remains largely intact.
That’s the core argument Michael B. Horn made before Congress this week. Testifying before the House Subcommittee on Higher Education and Workforce Development, Michael argued that AI’s limited impact on higher education isn’t a technology problem—it’s an operating model problem.
The institutions seeing the most promise aren’t the ones that have written the most AI policies; they’re the ones willing to rethink the underlying resources, processes, and priorities that have defined higher education for generations.
His testimony lays out a compelling roadmap for what that redesign could look like: reimagined assessments that raise the bar rather than lower it, real-world work embedded into the curriculum, and a new generation of AI-native institutions built from first principles rather than retrofitted from tradition.
If you care about whether higher education can actually deliver more value in an AI-driven economy—not just survive it—this is essential reading.
