By the end of 2025, only 19% of schools had a formal AI policy. Most districts that do have AI policies have built them as compliance exercises rather than governance instruments. A compliance policy tells people what they cannot do. A governance instrument tells an organization what it believes.
"Districts waiting for regulatory clarity before building governance architecture are waiting for something that may not arrive."
What a Defensible Policy Requires
- What do we believe AI should and should not do in our schools?
- How are those boundaries enforced at the classroom, school, and district level?
- What accountability structures exist when AI is used in ways that conflict with our policy?
- How are students, families, and staff informed about our AI practices and their rights?
- How does our policy evolve as AI systems and evidence change?
What Most District Policies Are Missing
The RAND Corporation's 2025 survey found that 80% of students report that teachers never explicitly taught them how to use AI for schoolwork. Most districts have built educator-facing governance and neglected the student-facing layer entirely. A complete AI policy architecture applies equal rigor to educators and students.
Sources: RAND Corporation (2025); U.S. DOL AI Literacy Framework (2026); FutureEd (March 2026); EO 14365 analysis.