Colleague AI report reveals early teacher adoption across 12 school districts
New data from 12 Washington districts shows how teachers use AI in class.
A new midyear implementation report from Colleague AI and the University of Washington offers one of the first systematic looks at generative AI use in K-12 classrooms. Covering 12 Washington State school districts from September 1 to December 31, 2025, the report aggregates platform data alongside district-provided administrative records. The districts vary widely — from small rural systems with a few thousand students to large urban districts serving up to 30,000 — giving the study broad demographic coverage. However, only a subset of districts provided complete administrative data, so findings linking teacher AI usage to student characteristics are labeled as preliminary signals.
Produced jointly by Colleague AI and the university's research team, the report focuses exclusively on how teachers interacted with the Colleague AI platform during the first half of the 2025-26 school year. While specific usage metrics are not detailed in the abstract, the report's value lies in its real-world, multi-district scope. It marks an early attempt to understand adoption patterns, potential equity issues, and the practical integration of generative AI tools into daily teaching workflows. The research, published on arXiv, reflects a growing movement to study AI in education beyond controlled lab settings and into live classrooms.
- Report covers 12 Washington State districts using Colleague AI from Sept-Dec 2025
- Districts range from a few thousand to 30,000 students across rural, suburban, and urban areas
- Findings linking teacher AI use to student characteristics are preliminary due to incomplete data from some districts
Why It Matters
First large-scale, real-world data on teacher AI adoption in K-12 classrooms reveals early patterns and gaps.