When it comes to AI policies in schools, there’s a spectrum. And it puts student learning and safety at the center.

Two of the three largest school districts in the country restricted student AI access within a day of each other this month. Los Angeles Unified paused generative AI on every student device. New York City banned it for 2-K through 8th grade while phasing in supervised access for high schoolers. Meanwhile, districts like Dublin Unified in California and Charleston County in South Carolina are already several steps further down the road, building tiered, age-based policies instead of a single on/off switch.

None of this is really about whether AI belongs in schools. It’s about sequencing: pause or restrict unmanaged use, build the visibility and criteria to manage it, then reopen access on better terms — usually calibrated by age. Wherever a district falls on that spectrum, the same requirement follows it: the ability to see what’s happening on its own network. (Lightspeed can help.)

Ban vs. Allow Is the Wrong Question (or Just the First One)

A ban with a review committee, a data plan, and a family-reporting requirement attached isn’t a rejection of AI — it’s sequencing. LAUSD board member Kelly Gonez made that point directly after her district’s pause: it isn’t “a pause for forever,” it’s the room to figure out what the policy should look like. NYC’s policy runs on the same logic, with a one-year moratorium for younger grades, a coalition studying impact, and a literacy-paired on-ramp for high schoolers.

Age is the axis that keeps surfacing. Amanda Bickerstaff, CEO of AI for Education, has argued for months that young students shouldn’t be handed open access to AI chatbots — not because AI has no place in the classroom, but because the tools demand judgment that younger students haven’t developed yet. LAUSD’s own prior policy reflected that instinct before this month’s pause: access at 13 and up, under supervision. NYC’s new policy draws the same line at the middle-to-high-school transition.

So the real question isn’t “ban or allow.” It’s:

• Where does AI help versus harm?
• What age groups should use AI, and under what supervision?
• What counts as an approved tool vs. an unsanctioned risk?
• How do we actually enforce our policies?

A policy is only as good as a district’s ability to enforce and measure it. That last question is where most of the real differentiation between districts is happening right now.

Where Districts Actually Land on the AI Spectrum

LAUSD	NYC	Charleston County	Dublin Unified	Catawba County
Ban, for now	Grade-based ban	Phased rollout	Tiered policies by age	AI literacy first
Paused generative AI on all ~378,000 student devices; a reversal of the prior 13-and-up, supervised-access policy	Bans student-facing AI for 2-K–8th grade; phases in supervised, literacy-paired access for high schoolers	Board-approved AI policy (Jan. 2026) built with AI for Education; training and literacy work rolling out through 2026–27; moving toward district-wide enforcement	Among the earlier districts to pass a formal AI resolution; framework explicitly calls for AI use to be age-specific and developmentally appropriate	Three years into a state-grant-funded rollout of GenAI tools across all 28 schools, paired with a district-adopted policy governing responsible use and literacy for staff and students
Learn more	Learn more	Learn more	Learn more	Learn more

Los Angeles: ban, for now. LAUSD’s interim CIO Douglas Le told LAist the district’s filtering software can also track student device usage in detail — data the district says will flow to families through its Parent Portal, teachers, and the board as the “what’s next” conversation continues.

New York City: similar posture, earlier in the buying cycle. NYC’s moratorium covers nearly 600,000 students and disables AI features across dozens of approved ed-tech programs — a heavier enforcement lift than a single-vendor block.

Dublin: tiered by design, not by pause. Rather than a blanket restriction, Dublin Unified built its approach around age-appropriate, developmentally grounded AI use from the start — a model closer to where LAUSD and NYC are both headed once their pause periods end.

Charleston County: policy first, enforcement next. Working with AI for Education since late 2025, Charleston County’s board approved a district-wide AI policy in January 2026 built around a “safe, ethical, effective” framework, with teacher training and literacy instruction rolling out through the 2026–27 school year. From screen time to AI, Charleston has announced use of the Lightspeed Parent Portal this year to give parents greater ed-tech visibility.

Catawba County: allow, with literacy built in from the start. Catawba County Schools is three years into a state-grant-funded partnership with AI for Education to roll out GenAI tools across all 28 of its schools, paired with a district-adopted technology policy that governs responsible AI use and literacy for both staff and students — the clearest “allow” example on this list, built on guardrails rather than restriction.

Other districts are already showing what governance in action looks like day to day. Barrow County Schools (GA) used Lightspeed Insight to catch students routing around district AI policy into public ChatGPT, then moved them into a sanctioned, monitored alternative. Oley Valley School District (PA) turned grade-level usage data into an evidence-based conversation with its school board — the same kind of reporting LAUSD says it’s building toward through its Parent Portal.

Coast to coast…from one end of the spectrum to another…Lightspeed helps districts enforce and monitor their AI policies.

The Common Thread: Age-Appropriate Governance & Guardrails

Whether a district is at “ban,” “semi-ban,” or “allow with guardrails,” the requirement underneath doesn’t change: someone has to be able to see what’s actually happening on student devices.

• Lightspeed Insight shows what AI tools are being used: which apps students are using, how much time they’re spending, and where usage drifts toward unsanctioned tools and shadow AI.
• Lightspeed Filter enforces whatever the current policy is — blocking AI categories entirely, or scoping access by age group, grade band, or approved tool list.
• Lightspeed Classroom gives teachers visibility into AI use in class with real-time notifications.
• Lightspeed Alert flags concerning patterns in real time, so a policy gap doesn’t turn into a safety gap.

That’s the connective tissue across every posture on this spectrum — a ban only holds if you can see whether it’s holding, and a tiered policy only works if you can set age-appropriate policies and monitor against them.

What Districts Should Do Now

1. Decide where you are on the spectrum today — full pause, grade-tiered, or open with guardrails — and write it down, even if it’s provisional.
2. Set the age lines explicitly. Match access to grade bands rather than leaving it to individual teachers or classrooms to decide.
3. Get visibility before you finalize the policy, not after. Districts that already have usage data are making better-informed calls than those building policy in the dark.
4. Plan the reporting loop — to families, teachers, and the board — as part of the policy, not as an afterthought once something goes wrong.
5. Revisit on a set schedule. Treat the current posture as a checkpoint, not a permanent answer.

Wherever Your District Falls, We Can Help You Get There