Three big districts wrote three different answers to the same question

The classroom note

Three of the largest school systems in the United States have now published AI policies, and reading them side by side shows how much genuine disagreement remains about what a classroom should allow. In Prince George's County, Maryland, Board Policy 0123, adopted 25 April 2024 and amended 26 June 2025, treats AI as a supplemental instructional tool subject to approval. In Philadelphia, the district's Digital Access Hub names two specific approved products by grade band. In New York City, the newest of the three, Chancellor Samuels's guidance for the 2026-27 school year bans student-facing generative AI outright below grade 9.

What the evidence says

The documents differ in structure as much as content. PGCPS's board policy is a governance document: it requires AI platforms to be vetted and approved by the Superintendent or designee, requires citation of AI use just as with any other source, requires parental consent before offering certain AI services to students, and prohibits AI replacing teaching by classroom teachers or submitting AI output as the original work. Philadelphia's page is a product list, not a governance framework: it names Google Gemini as approved for staff now and students from late January 2025, restricted to grades 9-12, and Adobe Express with Firefly as approved for all staff and students, K-12, immediately. New York's policy is the most restrictive of the three: student-facing generative AI is banned in grades 2K through 8, limited to approved, vetted programs in grades 9-12 depending on school and course, alongside separate 1:1 screen-time limits by grade band, with exceptions preserved for assistive technology under IEP or 504 plans.

The implementation question

None of the three documents is simply an AI policy in the same sense. PGCPS regulates use once a tool is approved; Philadelphia regulates which tools exist to be used at all; New York regulates access by age before either question arises. A family moving a child between these three districts would find not just different rules but a different starting question: what can I use, which tool can I use, or can I use anything yet. New York's policy also names a concrete accountability step the other two do not: a report on future policy recommendations is due in April 2027, meaning the current restrictions are explicitly framed as provisional rather than settled.

What holds and what fails

Comparing published policies holds as a way to see genuine, unresolved disagreement among serious school systems, not a single emerging consensus; a family or teacher reading only one district's policy could easily mistake local rules for a national norm. It fails as a way to judge which approach is correct, since none of the three documents cites outcome evidence for its own restrictions or permissions; each reflects a governance choice, not a measured result. That distinction matters more as more districts publish policies this year.

• Does the home district's policy regulate tool approval, product choice, or age-based access, and does that match what is actually happening in classrooms?
• Is the current policy explicitly provisional, like New York's, or does it read as a settled rule?
• What would change under each of the three approaches if a family or teacher disagreed with the restriction?

Three major districts, writing independently, reached three different governance models within roughly two years of each other, which is itself the clearest evidence that no settled answer yet exists.