ACTUAL Intelligence: What AI in Education Gets Right — and What It Can't Replace

Introduction

EDUCAUSE — the nonprofit that shapes technology strategy for higher education institutions — spent months in early 2026 developing a framework for what AI-enhanced education should actually look like. The resulting framework is built around an acronym: ACTUAL. Agency. Connection. Trust. Uniqueness. Adaptability. Lifelong Learning.

What strikes me about this framework is what it prioritizes. It doesn't start with capabilities. It starts with values. It asks not "what can AI do in education?" but "what is education actually for, and what role should AI play in serving that purpose?" That ordering matters.

What the Framework Says — and Doesn't Say

Agency: AI-enhanced education should increase learner agency — the ability to direct one's own learning, make meaningful choices, and develop genuine self-understanding. Connection: Learning is fundamentally relational. AI can provide personalized content and adaptive feedback, but it cannot replicate the human relationships through which the deepest learning happens. Trust: AI systems in education require trust from learners — trust that the technology is used ethically, that data is protected, that AI support doesn't undermine the integrity of assessment or credential. Uniqueness: Every learner is different. AI has genuine potential to personalize learning in ways traditional classroom instruction cannot. Adaptability: Learning doesn't end. Lifelong Learning: The goal of education is not credential acquisition but the development of human beings who can continue learning throughout their lives.

What AI Is Actually Getting Right in Education

Personalized learning pathways. AI tutoring systems that provide immediate, patient, responsive feedback — available at 2 AM when a student is stuck, without judgment. Language learning applications with high efficacy. Accessibility tools that create genuine inclusion for learners with disabilities. Early identification of students at risk of falling behind, enabling interventions before failure rather than after. These are genuine achievements.

What AI Cannot Replace

The teacher who sees a student struggling and knows, from relationship and context, that the struggle is about something beyond the material. The classroom conversation where ideas collide in ways that couldn't be scripted or predicted. The mentor whose belief in a student's potential changes what that student believes is possible. The experience of being witnessed — genuinely known — by another person who has invested in your development.

These are not sentimental attachments to an old model. They are descriptions of what the research consistently shows produces lasting learning, genuine development, and the kind of human growth that persists long after the course is over.

Closing Reflection

The best version of AI in education is one where the technology handles the tasks it's genuinely better at — personalization, availability, patience, pattern recognition — so that human teachers can focus on the work only they can do: building relationships, creating community, modeling intellectual life, mentoring development. That division of labor is possible. But it requires intentional design, not just technology deployment.

---

Sources

EDUCAUSE Review. ACTUAL Intelligence: Practitioner Perspectives on Centering the Human in the Age of AI. May 2026. https://er.educause.edu/articles/2026/5/actual-intelligence-practitioner-perspectives-on-centering-the-human-in-the-age-of-ai

Brookings Institution. AI's Future for Students Is in Our Hands. https://www.brookings.edu/articles/ais-future-for-students-is-in-our-hands/

PMC. AI in Personalized Learning for Older Adults and Intergenerational Engagement: A Systematic Review. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12436965/

ERIC / U.S. Department of Education. Modern Learning Strategies in the Age of Digital Technology. https://files.eric.ed.gov/fulltext/EJ1483551.pdf