pith. sign in

arxiv: 2605.16277 · v1 · pith:JRYU52ZQnew · submitted 2026-04-08 · 💻 cs.CY · cs.AI

Generative AI in K-12 Classrooms: A Midyear Implementation Report

Lief Esbenshade , Alex Liu , Michael Xiao , Zewei Tian , Min Sun , Zachary Zhang , Thomas Han , Yulia Lapicus
show 1 more author
Kevin He
This is my paper

Pith reviewed 2026-05-21 09:00 UTC · model grok-4.3

classification 💻 cs.CY cs.AI
keywords generative AIK-12 educationteacher AI useimplementation reportWashington State districtsclassroom technology adoptionmid-year data
0
0 comments X

The pith

Teachers in 12 Washington districts used Colleague AI for lesson planning and grading during the first half of the 2025-26 school year.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The report compiles platform usage logs from Colleague AI with administrative records supplied by a subset of participating districts to map how K-12 teachers actually interacted with generative AI tools. Districts range from small rural to large urban settings and vary in size from a few thousand to thirty thousand students. The resulting description supplies an early, data-grounded picture of adoption patterns rather than controlled experiments or long-term outcomes.

Core claim

Across the period September 1 to December 31 2025, teachers in the twelve districts logged measurable engagement with Colleague AI, and the aggregated records permit preliminary examination of whether usage levels relate to student demographic or performance characteristics.

What carries the argument

Joint aggregation of Colleague AI platform logs and district administrative records that link teacher activity to student characteristics.

If this is right

  • Usage patterns can serve as a baseline against which later adoption in the same districts can be compared.
  • Preliminary correlations between teacher AI use and student traits can guide targeted support or equity checks in participating districts.
  • The range of district sizes and locations included allows initial comparison of AI uptake across rural, suburban, and urban contexts.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the observed usage holds, districts may need to adjust professional development resources toward the specific tasks teachers already perform with the tool.
  • The data structure could support later studies that track whether AI use shifts over a full academic year or differs by subject area.

Load-bearing premise

The administrative data from the subset of districts willing and able to share mid-year records is representative enough to support preliminary signals about links between teacher AI use and student characteristics.

What would settle it

Full-year data or records from the remaining districts showing no consistent relationship between teacher AI usage and student characteristics, or showing usage patterns that differ sharply from the mid-year snapshot.

read the original abstract

This mid-year report summarizes teacher use of Colleague AI across 12 Washington State school districts from September 1 to December 31, 2025. Produced jointly by Colleague AI and AmplifyLearn.AI at the University of Washington, this report aggregates platform data and district-provided administrative records to provide an early look at how teachers engaged with AI during the first half of the 2025-26 school year. The districts vary in size from small districts with a few thousand students to large districts with up to thirty thousand students. The districts are rural, suburban, and urban. Only a subset of districts were able to provide mid-year administrative data, and findings that link teachers' use of Colleague AI to student characteristics should be interpreted as preliminary signals.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. This mid-year report summarizes teacher use of Colleague AI across 12 Washington State school districts from September 1 to December 31, 2025. It aggregates platform usage data and district-provided administrative records to offer an early descriptive overview of engagement patterns, noting that the districts range from small (a few thousand students) to large (up to thirty thousand) and include rural, suburban, and urban settings. Only a subset of districts supplied mid-year administrative data, and the report explicitly states that any findings linking teacher AI use to student characteristics should be treated as preliminary signals.

Significance. If the descriptive patterns hold after addressing data limitations, the report supplies timely, real-world evidence on generative AI adoption in K-12 settings during the first half of the 2025-26 school year. Its transparency in flagging the preliminary status of student-characteristic associations is a strength, as is the joint production by Colleague AI and AmplifyLearn.AI at the University of Washington, which may facilitate future reproducible analyses if raw aggregates or code are shared.

major comments (1)
  1. [Abstract] Abstract: The report correctly labels links between teacher Colleague AI use and student characteristics as preliminary because only a subset of districts provided administrative records. However, it provides no comparison of district size, urbanicity, or IT/data-system maturity between the sharing subset and the full set of 12 districts. Without such a comparison, the preliminary signals cannot be evaluated for selection bias, which directly affects interpretability of the central descriptive claim.
minor comments (1)
  1. [Abstract] The abstract states that districts vary in size and setting but does not supply even summary counts or a table of district characteristics; adding this would improve clarity without altering the preliminary framing.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive review and for recognizing the report's timeliness and transparency regarding preliminary findings. We address the single major comment below and outline the revisions we will make.

read point-by-point responses

  1. Referee: The report correctly labels links between teacher Colleague AI use and student characteristics as preliminary because only a subset of districts provided administrative records. However, it provides no comparison of district size, urbanicity, or IT/data-system maturity between the sharing subset and the full set of 12 districts. Without such a comparison, the preliminary signals cannot be evaluated for selection bias, which directly affects interpretability of the central descriptive claim.

    Authors: We agree that a comparison of the districts that supplied administrative records against the full set of 12 would allow readers to better evaluate selection bias. The current manuscript does not include this analysis. In the revised version we will add a concise descriptive paragraph and, if space permits, a small table comparing the two groups on district size (student enrollment) and urbanicity (rural/suburban/urban classification) using the information already available to us. We lack systematic measures of IT or data-system maturity across all districts and therefore cannot supply a full comparison on that dimension; we will instead note this explicitly as an additional limitation. These changes increase transparency without overstating what the data support. revision: partial

Circularity Check

0 steps flagged

No circularity: purely descriptive data aggregation report

full rationale

The document is a mid-year implementation report summarizing observed teacher use of Colleague AI across 12 Washington districts using platform logs and a subset of administrative records. It contains no equations, derivations, fitted parameters, predictions, or first-principles claims. All content is descriptive aggregation with explicit caveats labeling linked findings as preliminary signals due to the subset limitation. No load-bearing step reduces to a self-definition, self-citation chain, or renamed input; the report is self-contained as an early observational summary without any claimed mathematical or predictive structure.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The report rests on the assumption that platform logs and partial administrative records accurately capture teacher engagement without detailed validation or bias checks described.

axioms (1)
  • domain assumption Platform usage logs and district administrative records accurately reflect actual teacher engagement with Colleague AI.
    Invoked when aggregating data to describe engagement patterns; no validation steps mentioned.

pith-pipeline@v0.9.0 · 5679 in / 1130 out tokens · 54499 ms · 2026-05-21T09:00:59.517528+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

  1. [1]

    Campuzano, L., Dynarski, M., Agodini, R., & Rall, K. (2009). Effectiveness of Reading and Mathematics Software Products: Findings From Two Student Cohorts. NCEE 2009-4041. National Center for Education Evaluation and Regional Assistance. Esbenshade, L., Sarkar, S., Nucci, D., Edwards, A., Nielsen, S., Rosenberg, J. M., … & He, K. (2025). Emerging Patterns...

    work page arXiv 2009

  2. [2]

    Turgut, R., Sahin, S., & Huerta, M. (2016). Teachers’ perceptions of effective school-wide programs and strategies for English language learners. Learning Environments Research , 19(2), 175–194. https://doi.org/10.1007/s10984-015-9173-6 References 27 Appendix: Report Data Sources The AmplifyLearn.AI research team would like to thank the districts that pro...

    work page doi:10.1007/s10984-015-9173-6 2016