We were not waiting for the state to tell us what was allowed. We were deciding what we believed, and then we were ready for whatever the state said next.
Twenty-seven states are now regulating how artificial intelligence gets used in classrooms, and no two of them agree on what that means. FutureEd is tracking 77 separate bills this session addressing AI in instruction alone, part of a broader wave of nearly 100 state bills and more than 1,500 AI-related proposals introduced nationwide (DiMarco, 2026). New York is moving to ban most classroom AI use below ninth grade. Alabama is requiring every high schooler to complete a computer science course built around it. Hawaii wants a task force to study the question before anyone commits to an answer. Thirty-five states and Puerto Rico now have some form of official guidance on the books, most of it written in the last two years (Boser, 2026). That is not a shared understanding of what AI should do in a classroom. That is 35 different understandings, each one convinced it got there first.
CONSTRUCTED ANECDOTE, CONFIRMED BY LUIS. Ready for publication.
Early in my superintendency, we faced a similar gap, waiting for state guidance that had not caught up to what was already happening in our classrooms. I remember sitting with my cabinet and deciding we would not wait. We wrote our own principle first: technology serves the teacher's judgment, it does not replace it. When the state guidance finally arrived two years later, we found we had already been living inside something close to it. We had not guessed correctly. We had simply decided what we believed before someone else decided it for us.
That is the choice in front of every district leader watching this legislative session unfold.
What the Sprawl Actually Looks Like
No single state's approach is wrong on its own terms. New York's caution reflects real concerns about developmental readiness. Alabama's coursework mandate reflects a real belief that AI literacy belongs alongside math and reading. Oklahoma's Responsible Technology in Schools Act, which bars AI from serving as the primary basis for grading, discipline, or placement decisions, reflects a real concern about human judgment getting automated out of the room (DiMarco, 2026). Each state is solving for something legitimate.
But a district superintendent does not operate inside one state's logic. She operates inside her own building, in front of her own teachers, on a Tuesday morning when a ninth grader submits an essay that reads a little too polished. The state legislature is not in that room. The policy the district wrote, or failed to write, is the only thing that is.
Two Districts, Two Philosophies
Ohio offers a clean comparison, because Ohio forced the question. Every district in the state had to adopt a formal AI policy by July 1, 2026 (GovTech, 2026). Columbus City Schools answered with a short, plain policy: AI is a supplement, not a substitute for student effort or teacher judgment, teachers decide assignment by assignment whether it is allowed, submitting AI-generated work as one's own falls under the district's code of conduct, and only district-approved tools may be used at all (WOSU, 2026).
New York City answered differently: a Central AI Task Force, an external AI Advisory Council, and a Data Privacy Working Group, with a vetting process banded across K-5, 6-8, and 9-12 grade spans before any tool reaches a classroom (NYC Public Schools, 2026). That is not a criticism. A system serving roughly a million students needs infrastructure a district a fraction of that size does not. But size is not the variable that determines whether a policy holds together. Clarity is.
Coherence Is a Choice, Not a Compliance Deadline
I have made this argument before technology forced the question: when leaders adopt systems thinking, they do not simply react to what arrives, they build a structure sturdy enough to absorb it (Valentino, 2025). Columbus did not have better information than any other Ohio district facing the same July deadline. It had a clearer internal principle, decided in advance, that the legislation could then attach itself to instead of dictate from scratch.
That is the real dividing line this legislative session is drawing, not between states that regulate AI and states that don't, but between districts that had already decided what they believed and districts that are still waiting to be told.
The Choice in Front of You
You do not need to wait for your state to finish arguing with itself. You need one page and one principle: what does AI do in your classrooms, and what does it never do. Write that down before the legislature writes it for you. The districts that wait will spend this year in compliance meetings. The districts that already decided will spend it teaching.
References
Boser, U. (2026, January 20). Where states err in their AI education guidance, and how to fix it. Forbes. https://www.forbes.com/sites/ulrichboser/2026/01/20/where-states-err-in-their-ai-education-guidance--and-how-to-fix-it/
DiMarco, B. (2026, July 13). Legislative tracker: 2026 state AI in education bills. FutureEd. https://www.future-ed.org/legislative-tracker-2026-state-ai-in-education-bills/
GovTech. (2026, January 9). Ohio unveils model AI policy for use by K-12 schools. https://www.govtech.com/education/k-12/ohio-unveils-model-ai-policy-for-use-by-k-12-schools
NYC Public Schools. (2026, March 24). Guidance on artificial intelligence. https://www.schools.nyc.gov/about-us/vision-and-mission/artificial-intelligence/guidance-on-artificial-intelligence-full
Valentino, L. R. (2025). Thought leadership in education: A comprehensive exploration of transformative educational ideas. Valgar LLC.
WOSU. (2026, March 31). Columbus City Schools to consider new AI policy giving teachers discretion on student use. https://www.wosu.org/politics-government/2026-03-31/columbus-city-schools-to-consider-new-ai-policy-giving-teachers-discretion-on-student-use
Let's Talk
|
Every district in America is one page away from a real AI policy, and most of them don't know it yet. THIS WEEK Pick a moment this week to write down, in one page, what AI is allowed to do in your classrooms and what it is never allowed to do. Don't wait for a committee. Draft it alone first, in plain language, the way you would explain it to a teacher who has never given AI a second thought. Then take it to three people: a teacher who uses AI often, a teacher who does not trust it, and someone outside education entirely. Ask each of them what is missing. You may find your first draft holds up. You may find it needs real work. Either answer is useful. The page is what matters, not that it is perfect on the first try. Write it. Send it to your cabinet. Then wait for the legislature to catch up to you instead of the other way around. Please share in the comments. I will respond. |
DISTRICT LEADER PODCAST
tarting with this issue, the podcast broadens its lens. Alongside conversations with superintendents and principals, we will feature non-traditional education leaders whose work outside the K-12 system still shapes how we lead. Beth Rudden is the first of these voices, a former archaeologist who now leads Bast AI after running a $34 billion division as Chief Data Officer at IBM Managed Services. In this conversation with Mike Palmer, she explains how Bast AI checks its answers against real data instead of letting a model guess and hope nobody notices, work with real stakes for any classroom trusting AI.
“Going through the process of math, is about the learning how to take things in the real world, and then in the mathematical world, and again in the real world.” - Beth Rudden
EDUPRENEURS NETWORK • DEEP DIVE
Value-Based Procurement in Education: Moving Beyond the Lowest Bid
This week's issue makes the case that districts should decide their AI principle before the state hands them one. Edupreneurs Network makes the same argument from the procurement side. Read the full piece at edupreneurs.network. Most districts still default to the lowest bid, then discover the real cost shows up later in implementation headaches and weak outcomes. Value-based procurement asks a harder question up front, what does this provider actually deliver, and does it match what we believe good instruction looks like. Coherence and value-based decision-making are the same habit of mind. Decide your standard first. Let compliance catch up to you.
From the Bookshelf - Thought Leadership
Systems Thinking and Organizational Tansformation
From Chapter 4, "Creating Policy That Doesn't Fight Itself," in Thought Leadership in Education, page 84:
"One of the most common challenges of education is conflicting policies. A state might push for one approach while a district enforces something else entirely. The result is confusion and compliance over innovation." I wrote that a year before this legislative session produced seventy-seven competing state bills on AI alone. The book goes deeper into what policy alignment actually requires across local, state, and federal lines, and why misalignment quietly undermines reform long before anyone notices.
Additional Resources
|
WEBSITE |
Legislative Tracker: 2026 State AI in Education Bills FutureEd's full bill-by-bill tracker behind this week's numbers, 77 bills across 27 states addressing AI in classroom instruction, updated throughout the 2026 legislative session. Explore → |
|
ARTICLE |
Columbus City Schools Approves New AI Policy The Columbus story in full, WOSU's on-the-ground reporting on how Ohio's largest district built its teacher-discretion AI policy ahead of the state's July 2026 deadline. Read more → |
|
ARTICLE |
Ohio Unveils Model AI Policy for Use by K-12 Schools GovTech on the state mandate and model policy language behind Ohio's July 1 deadline, the framework every district in the state, including Columbus, had to answer to. Read more → |
Stop typing what you could say in 10 seconds.
Wispr Flow turns your voice into clean, professional text inside any app. Emails, Slack, client updates — speak once, send without editing. 4x faster than typing.