Blog — Policy Watch

Who's getting left behind, and how

Issue No. 1 — dated updates on the decisions being made about grassroots families, not just the decisions grassroots families are making.

Last verified: July 2026

The data privacy exposure nobody signed up for

Most AI tools reaching classrooms weren't built with K-12 student-data law in mind. FERPA and COPPA exposure is real the moment a tool built for adult users gets handed to a minor — and it's the top fear administrators name, ahead of cheating.

The free-to-paid trap

Many AI tools currently free to schools are free because they're building market share, not out of generosity. Once a district's curriculum depends on a tool, the leverage shifts. Watch for contract terms before adoption, not after — see the Administrator checklist in the Playbook.

The equity gap, in numbers

Private schools report roughly 18% daily AI use versus 11% in public schools. Teachers at private schools are more than twice as likely to have received formal AI training — 45% versus 21%. Higher-poverty public districts are the least likely of any group to offer structured AI learning for staff or students. Left unmanaged, AI doesn't level the playing field. It widens it.

Classical & Christian schools: not the holdouts you'd expect

In the 2024 ACSI/Cardus survey, roughly 38% of Christian-school educators used AI at least sometimes, a nearly identical 37% never had, and only 18% reported an outright ban. The dominant story here isn't resistance — it's an active, open argument about where the lines should go, with top concerns clustering around cheating and integrity, critical-thinking effects, safety and privacy, and faith formation.

Detector accuracy: unreliable enough to matter

Independent testing of AI-detection tools shows accuracy ranging from the mid-50s to upper-90s percent, depending on the tool and the text. One widely cited study found detectors misclassified non-native English writers' essays as AI-generated at more than triple the rate of native speakers' — meaning the students least equipped to fight a wrong accusation are the ones most likely to receive one. A tool this inconsistent should never be the sole basis for a cheating accusation.

About this series

This series tracks decisions and numbers that shift — vendor terms, adoption gaps, detector reliability — and gets a fresh verification pass each issue rather than sitting as a static claim. Check the blog index for a newer issue before citing this one.

Comments are coming to this post.
We're setting up moderated, hold-for-approval commenting — no account required to post, everything reviewed before it appears. Check back soon.