Finkelstein's new framework guides AI in education with learning science principles
A principled approach ensures AI augments, not displaces, human learning in higher ed.
In a new paper published on arXiv (Oct 2025, revised Jul 2026), University of Colorado researcher Noah D. Finkelstein presents a principled framework for using AI in education, moving beyond the typical debate of promise versus peril or quick implementation. Drawing on decades of scholarship in learning sciences and educational technology, the paper outlines a set of principles that connect broad educational goals to concrete practices for curriculum design, instruction, assessment, and community building.
The framework clarifies the distinct roles of educators, learners, and technologies, ensuring that AI serves as an augmentation rather than a replacement for human capacities. It emphasizes preserving higher education's core missions: advancing meaningful learning, supporting democratic societies, and preparing students for dynamic futures. Finkelstein notes that the principles, initially shared in July 2025, remain relevant a year later, making this a practical, enduring guide for educators and policy makers navigating AI integration.
- Framework based on learning sciences research, not hype or fear, to guide AI integration
- Provides actionable practices for curriculum design, instruction, assessment, and community building
- Aims to augment human capacities and preserve core educational values like democratic engagement
Why It Matters
Helps educators and policymakers use AI to enhance learning without losing sight of human-centered goals.