Banyan Global Learning
AI Policy hubFree checklist

District AI Policy Checklist

Twenty-six questions your AI policy — and the governance around it — needs to answer. Organized by the five layers, because a policy that only covers compliance fails everywhere else. Bring it to your next cabinet meeting and mark what you can honestly check.

Layer 1 · Strategic Clarity

Before policy language: does the district know what it believes?

  • A board-approved statement of AI intent exists — one page, plain language, dated within the last 18 months.
  • The superintendent and cabinet can state the district’s AI posture in two minutes without a script.
  • AI strategy lives inside the district strategic plan, not in a separate binder nobody opens.
  • Success metrics are defined for both efficiency (staff time, cost) and learning quality — not just one.
  • A parent or reporter can find the district’s AI position on the public website in under a minute.

Layer 2 · Risk & Compliance

The layer most policies cover — but usually only halfway.

  • An inventory of every AI tool in district use exists and has been updated in the last 90 days.
  • Vendor evaluation includes AI-specific criteria: training data, bias, data flow, transparency.
  • The acceptable use policy names AI explicitly and was revised within the last 12 months.
  • FERPA and COPPA implications of each approved tool have been reviewed and documented.
  • Staff know the incident protocol for a data breach, a biased output, or an integrity violation — by name, not by guess.
  • Academic-integrity boundaries for AI are differentiated by grade band, not one rule K–12.

Layer 3 · Instructional Integration

The cognitive core — the layer that determines whether AI helps or hollows out learning.

  • The district has articulated which cognitive work must remain human-performed for learning to occur.
  • High-stakes assessments have been reviewed against what AI can now do, and updated where they broke.
  • Teachers have a shared framework for deciding when AI assists learning versus when it bypasses it.
  • Professional development covers pedagogy, ethics, and cognitive impact — not just tool walkthroughs.
  • Students are taught to evaluate AI outputs: what to trust, what to verify, what to reject.

Layer 4 · Operational Implementation

The tissue between the policy document and Monday morning.

  • A 6–12 month AI roadmap exists with milestones and named owners.
  • PD is sequenced over time and differentiated by role — not a one-and-done workshop.
  • A proactive parent and community engagement plan exists (not just reactive FAQs).
  • Students have a structured voice in AI governance decisions that affect them.
  • Building leaders apply consistent expectations across schools — checked, not assumed.

Layer 5 · Infrastructure & Future-Readiness

The decisions that quietly lock a district in — or keep it free.

  • A decision framework exists for cloud versus local/private AI solutions.
  • Infrastructure choices trace to educational principles, not vendor sales cycles.
  • Procurement criteria require interoperability and data portability.
  • Someone owns monitoring what’s next: agentic AI, multimodal models, AI companions.
  • A 2–3 year evolution roadmap exists that assumes the technology will keep moving.

How many could you honestly check?

The checklist tells you what to look at. The free 5-Layer Readiness Snapshot tells you where you stand — a 10-minute leadership self-assessment that returns a maturity rating per layer and your three most urgent priorities.

Take the free Snapshot