By Brett Roer, Founder & CEO of Amplify & Elevate Innovation
Earlier this month Candice Sears and I were in Westerville, Ohio, at the OESCA fall conference, talking with ESC and district leaders about implementing an AI Community Playbook. Every district in the room had been required by law to adopt an AI policy by July 1. About twenty minutes in, a superintendent near the back put the question that the whole room had been circling.
“We have the policy. What’s the strategy?”
I have heard a version of that question in Texas, in Pennsylvania, in New York, and in Indiana this year. The policy tells people what is allowed. It does not tell a district what it is trying to do, who gets to decide, or what a good year looks like. That is what an AI strategy for school districts is for, and in my experience it comes down to five decisions. The document is what you write once the decisions are made.
Decision one: what AI is for in your district
The first decision is purpose, and it is the one most districts skip because it feels obvious. It is not obvious. Ask five people in the same central office what AI is for and you will get five answers: efficiency, personalization, workforce readiness, equity, keeping up.
The districts I partner with settle this by listening before they decide. Students, families, educators, and staff, in structured conversations, saying what they hope AI gets right for their community and what worries them most. When the purpose statement comes out of that process, it is short, it is in plain language, and people recognize themselves in it. When it comes out of a cabinet meeting alone, it reads like every other district’s.
RAND’s survey panels put a number on why this matters now. In 2025, 54 percent of students and 53 percent of English language arts, math, and science teachers said they used AI for school, up more than 15 percentage points in one to two years. Only 35 percent of district leaders said they provided students with any training on it. The full RAND report is worth an hour of a cabinet’s time. The use is already in the building. The purpose is what the district gets to choose.
Decision two: who decides
The second decision is governance, and I want to be specific about what I mean. Not who signs the policy. Who is at the table when the next question comes up, because there is always a next question.
I built my work around AI advisory committees that deliberately bring together students, families, educators, leaders, and technical staff. I watched one of those committees work through a question about detection tools last fall. A student named the trust problem. A privacy officer explained the regulatory side. A teacher described the assessment design challenge. A parent named the anxiety she had been carrying since September. None of them alone could see the whole picture. Together they could, and the guidance they produced held, because they built it.
The U.S. Department of Education’s Dear Colleague letter of July 22, 2025 asks for exactly this. Among its five principles for AI in education is that stakeholders, and parents in particular, should participate meaningfully in decisions about adopting new technologies. A standing table is how a district does that without reinventing the process every time.
Decision three: which tools, and on what terms
This is where an EdTech AI strategy lives, and it is the decision districts tend to make backwards. The tool arrives, a teacher loves it, a purchase order follows, and the terms get worked out later.
I keep coming back to something Dr. Chad Stevens of ClassDojo told me on AmpED to 11. His team built ClassDojo for Districts with a customer advisory board of communication, academic, and technology leaders, and nearly everything those leaders asked for is on the roadmap. So the question I now put to every vendor a district is considering is simple. Who shaped your roadmap? If the answer is not educators, families, and leaders, that tells you something about whose problems the tool was built to solve.
The terms matter as much as the tool. A data processing agreement in place before anyone logs in. A clear answer on what student data is collected and where it goes. A named person in the district responsible for reviewing the tool a year from now, because AI products change faster than the contracts around them. If your district does not yet have a vetting process, this guide to vetting AI tools for schools is the place to start, and the roadmap piece from earlier this month shows where vetting sits in a three-year sequence.
Decision four: how people learn
The fourth decision is professional learning. Every district offers some. The strategy question is what kind.
Most AI professional development I have seen arrives as compliance: a mandatory session, a list of tools, a box checked. Teachers leave with a login and no reason to use it. The professional learning that works in the districts I partner with feels like discovery. It starts with the expertise teachers already have, gives them a mental model that keeps their judgment at the center, and lets them find their own meaningful task first. That is why I built the AI Sandwich Strategy™: human first, AI in the middle, human last. It gives a teacher a way to try something on Tuesday without handing over the part of the work that is theirs.
The strategy decision is to fund and schedule that kind of learning as a sustained cycle rather than a one-off event, and to differentiate it. A kindergarten teacher, a high school chemistry teacher, and a special education coordinator do not need the same session. The AI Professional Development work I do with districts is built around that difference.
Decision five: how leaders use AI themselves
The fifth decision is the one nobody puts in the strategy document, and it is the one that determines whether the other four are believed. Do the leaders use it?
I am not talking about a superintendent demonstrating a chatbot at a board meeting. I am talking about the AI tools for school leadership that quietly change where the hours go. The lens I use with principals is meaningful but menial: the work that has to be done, has to be done well, and does not require you specifically to be the one doing it. Drafting the first version of a staff memo from three bullet points. Turning a walkthrough note into coaching feedback the teacher can act on. Summarizing a board policy into what it means for your building on Monday. Converting one schedule change into the four different messages it requires for staff, students, families, and the district office.
None of that is the job. All of it eats the job. When a principal takes those tasks off the plate with AI and spends the recovered hour in a classroom or with a family, the staff notice, and the strategy stops being a document and starts being a culture. I wrote about where the hours actually go earlier this summer, and the leaders I coach usually find six to ten of these tasks in their first week of looking.
Writing it down, once the decisions are made
When a district has made those five decisions, the strategy document almost writes itself, and it is short. A purpose statement the community recognizes. A standing advisory table with named seats. A tool vetting and review process with someone’s name on it. A professional learning cycle with dates. A leadership commitment that is visible in how the cabinet actually works.
If your district is in Ohio and adopted the state model policy to meet the July deadline, this is the work that turns compliance into direction. The Ohio School Boards Association’s summary of House Bill 96 is clear that the policy has to address student and staff use. It says nothing about strategy, because strategy is yours to decide.
Back in Westerville, after the session, the superintendent who asked the question stayed behind. She said the thing she wanted most was not another framework. It was a room where her people could answer the five questions together, out loud, and leave owning the answers.
Which of the five has your district decided? And which one is still being decided by default?
Frequently asked questions
What should an AI strategy for school districts include?
An AI strategy for school districts should settle five things: what AI is for in the district (a purpose statement the community helped write), who decides ongoing questions (a standing advisory table with students, families, educators, leaders, and technical staff), which tools are used and on what terms (a vetting process, data agreements, and an annual review), how people learn (sustained, differentiated professional development), and how leaders use AI themselves. The written strategy is short once those decisions are made.
How is an AI strategy different from an AI policy?
An AI policy sets the rules: what is permitted, what is not, and how student data is protected. An AI strategy sets the direction: what the district is trying to achieve with AI, who is responsible for the next decision, and what progress looks like over two to three years. States such as Ohio now require a policy. None I have worked in requires a strategy, which is why so many districts have the first and not the second.
What AI tools should school leaders use?
The most useful AI tools for school leadership are the ones that take meaningful but menial tasks off a leader’s plate: drafting first versions of memos and family communications, turning walkthrough notes into coaching feedback, summarizing policy into building-level guidance, and producing the several versions of one message that different audiences need. The tool matters less than the habit of identifying which tasks qualify and spending the recovered time with people.
How do you build an EdTech AI strategy for a district?
Start with the purpose decision, so tools are chosen for a reason rather than because they arrived. Then put a vetting process in place that asks every vendor who shaped their roadmap, what student data the tool collects, and who in the district will review it in a year. Sign data processing agreements before anyone logs in. Pilot a small number of tools against success criteria set in advance, and scale only what earns its place.
A thought partner for the five decisions
Most districts can make these decisions themselves. What they usually want is someone who has sat in the room while other districts made them, and who can hold the process so leadership can take part in it rather than run it. That is the AI Strategic Advisory work I do at the superintendent level, and the Community AI Playbook is the listening process behind decisions one and two.
If you are past the policy and working out the strategy, I would welcome the conversation: brett@aeieducation.org, or book a conversation.
