Don’t Just Police AI, Start Leading It: 5 Things Every School Leader Should Be Doing With AI Right Now

I've been in public education for over 30 years, and if there's one thing I've learned, it's that "techy" has never meant what people think it means. It doesn't mean you code. It doesn't mean you understand how a server works. It just means: you're willing to click on things. That's it. That's the whole secret.

I've been clicking on things my whole career. As a language arts teacher I embraced word processing, on-line research, and computers in classrooms early on. I was part of the founding team that opened Research Triangle High School, an independent public charter school and North Carolina's first fully flipped high school, back when "flipped classroom" was a phrase most people had never heard.  I even co-authored a book about it: Flipping Your Classroom without Flipping Out, and have frequently led NCCAT seminars on smart ways to adapt new approaches to technology use in school to our classrooms. 

None of that makes me a technologist. I’m still not exactly sure what a server does or how the “cloud” works, but it does makes me someone who has been willing, again and again, to lean toward the new tool instead of away from it, try it out loud where my staff and students could see me fumble through it, and figure out what it's actually good for before deciding what to fear about it.

So when I tell you that AI is not the end of education as we know it, I'm not saying it as a cheerleader who hasn't thought it through. I'm saying it as someone who has now watched this exact movie play out several times.

Remember when calculators showed up and everyone said kids would never learn math again? Cliff Notes? They’ll never read a book again! Spell check? They’ll never learn to write! Copy & Paste meant cheating would be rampant and undetectable. Remember when shared documents meant group work would turn into "one kid does it, everyone else's name goes on it"?

Every single one of those tools followed the same pattern. The tool arrives. Communities bemoan lost skills. Students get ahold of it and try to use it to cheat or circumvent the work, because it’s the nature of kids to test limits. And then, because good teachers are relentless, teaching adapts. It always adapts. 

AI is not a break from that pattern. It's the next chapter in it. And the leaders who understand that they are living inside a familiar cycle, rather than a brand new crisis, are the ones who will be able to guide their staff and students through it with confidence instead of fear.

That's really the core of what I want to say to you as a fellow charter school leader: this is a paradigm shift, not a threat to be contained. The schools that get ahead of it, that lean in and learn to work with AI instead of spending all their energy trying to police and prevent it, are going to be the ones whose teachers and students come out the other side stronger. This is a moment to lead! Here are five places where I'd start.

1. Start With Yourself

Before you ask a single teacher to change anything, spend real time using AI on your own workload. This matters for two reasons. First, you cannot lead people credibly through a shift you haven't lived yourself. Second, and just as important, your own job is genuinely full of tasks AI can lighten.

Explore the different tools and study up on what makes them different. ChatGPT is easy and quickly becoming the generic go-to, but the way it sources information may not make it your most reliable or ethical tool for each task. Pick some different tools, and be willing to just, “click on stuff!” 

Try using an AI tool to draft your family newsletter, your board update, your first pass at a staff memo. Ask it to help you think through a tricky parent email, summarize a long policy document before a board meeting, or organize your action items after a leadership team meeting. Let it help you build a rubric, a survey, a first draft of a schedule. Create a cross walk of curriculum documents to help with vertical alignment. None of this replaces your judgment. It replaces the blank page and the tedious first draft, which frees up your time and mental energy for the parts of the job that actually need you: the relationships, the hard conversations, the vision-setting.

Choose a few tools to become your go-to. BC many AI tools are generative, the more you use and give feedback on final products, the better it will learn your style and be useful to you. 

When your staff sees you using AI openly, imperfectly, and usefully, you've already done something important. You've modeled that this is a tool you pick up and learn, not a threat you brace against. And you'll have firsthand, credible knowledge of what it does well and where it falls short, which you'll need for everything that follows.

2. Teach Your Staff to Use It for Their Own Workflows

Consider bite-sized PD topics, real-world tasks where you can give a new AI "appetizer" at the start of PLCs or PD sessions. Your early adapters will catch right on, and then their peers will start to peek over their shoulders and catch their enthusiasm. 

Once you've found your own footing, turn outward. Your teachers are drowning in the same kinds of tasks you are: lesson planning, differentiation, IEP accommodations, parent communication, grading feedback, unit design. AI can genuinely lighten that load, but many teachers won't discover that on their own. They need you to show them, give them permission, and give them time to practice.

Build in real professional development time, not a single one-off session, for teachers to explore AI as a planning partner. Show them how to ask for three tiers of a reading passage instead of writing all three from scratch. Show them how to generate a first draft of parent communication in another language, or a differentiated worksheet for a student who needs more scaffolding, or discussion questions pitched at different levels of Bloom's Taxonomy for the same text.

This is also where the recommended teacher tools from my own recent AI training come in handy. Gemini's focus mode inside Google Classroom can streamline how students interact with AI so it's giving targeted academic support instead of just handing out answers. NotebookLM lets teachers build a repository of their own curated source material so AI responses stay grounded in what the teacher actually wants students working from, instead of the wide open internet. Tools like these help teachers use AI for guidance rather than generation, which is the distinction that matters most.

The goal isn't to make your teachers dependent on AI, it’s to give them back time and energy so their best thinking goes toward the parts of teaching that only a human can do: building relationships, reading the room, and designing the kind of learning experience a machine simply can't replicate.

3. Build Clear, Living Guidelines, Not Just Rules Change the Pedagogy

Safety First: there are some real concerns and horror stories about student data, student exposure to bad advice, chatbots providing harmful guidance, bias in materials, unethical use of AI tools. Yes, it’s imperative to set up protections and clear guidelines and expectations now. 

At the same time, here's the part where I'll ask you to resist what is often our  instinct as a school leader. Our first move, when something new and slightly scary shows up, is often to write a policy that says no, or that tries to catch and punish misuse. I understand the impulse. But banning AI outright is unenforceable, and leaning hard into detection tools is genuinely risky.

This past May, a student at an NC high school was falsely accused of cheating after a teacher relied on three unreliable AI detection tools. She was only cleared when another teacher reviewed her document's version history. That student then went on to launch a petition and speak in front of her district's board of education, asking for the kind of clear policy and appeals process her district didn't yet have. That story should sit with all of us. AI detectors carry high false-positive rates, and handwritten-only mandates create real accessibility barriers for kids who rely on assistive technology. Reactive policing isn't just ineffective; it can actively harm the students we're trying to protect.

So build guidelines instead of just rules, and build them at both the whole-school level and the classroom level. At the whole-school level, be transparent: tell students and families what AI use looks like across your building, what's expected, and what support is available if something goes wrong. At the classroom level, give teachers room to set expectations specific to their content and their assignments, because AI use in a coding class looks different than AI use in a creative writing class or an art program.

Shift the whole culture toward transparency and process instead of surveillance. Here are a few ideas I’m encouraging my own staff to use that are worth building into your school's approach:

• The Breadcrumb Method. Give real weight, in grading and in classroom time, to brainstorm maps, outlines, and drafts. When the process is part of the curriculum, the product stops being the only thing that counts.
• Living Documents. Have students do sustained work in shared spaces like Google Docs so teachers can see the evolution of ideas in real time. This isn't spying. It's coaching, the same way we'd want to sit beside a student and watch them think.
• AI Audit Trails. When students do use AI, ask them for a short process memo: what prompts did you use, how did you verify what came back, what did you change and why. That single habit builds both accountability and metacognition at the same time.
• New assessment models. A five minute oral defense of an argument's structure. Grading the evolution of a concept map over a week. Assessing a live debate or a group's real-time problem solving. These all measure something AI can't produce for a student: the thinking itself.

4. Help Teachers Teach Students How to Use AI Well

This is the piece I'd encourage every principal to treat as non-negotiable. AI literacy is a skill now, the same category as study skills or social-emotional skills, and it belongs embedded across every subject, not tucked into a single computer science unit once a year. Our students are going to grow up in a world full of AI. They need our guidance now to become safe, ethical, and genuinely strong users of it, not despite our teaching but because of it.

The heart of this is a simple but important reframe: use AI to build critical thinking, not bypass it. Here's the encouraging truth, straight from what the ISTE Profile of an AI-Ready Graduate and Bloom's Taxonomy both point us toward: there are things AI simply cannot do that students still can. Students can analyze, breaking a problem or text apart and explaining why something works. They can evaluate, judging AI's own output and deciding which sources or solutions to trust. They can create something that reflects their own voice, culture, and lived experience, since AI recombines patterns but has no experiences of its own to draw from. They can apply ideas in messy, real, hands-on situations AI can't physically enter. And they alone can choose their own goals and values, taking ownership of their work and making ethical choices about when and how to use a tool like AI at all.

Think of it as a flipped Bloom's Taxonomy for the AI era. Let AI support the foundational, lower-order tasks: retrieving facts, summarizing, drafting a first pass. Then let good teaching and human guidance take over for the higher-order work: analyzing, evaluating, and creating. AI and students become partners in learning, with AI handling the routine so students and teachers can focus on what matters most. With my team, we work on “assignment makeovers,” simple techniques that can help shift that focus, that ask for prompts that are specific to the learner, that call for application and synthesis rather than simple retrieval, and that build in transparency, so students are charting how they used AI and how their own thinking shifted along the way.

5. Bring Families Into the Conversation

Here's a piece I'd argue is just as important as the first four, and it's the one that's easiest to skip: lead your family and school community’s journey into AI, too. We spend so much energy getting our own staff and students ready that it's tempting to let parents catch up on their own. But our families are trying to parent kids through something they didn't grow up with, using instincts built for a different world, and they need our help the same way our teachers do.

Think about it the way I explained it to our own Exploris families at a recent parent workshop: understanding AI is a lot like learning to drive a car. You need safe operating rules and the situational awareness to react when something unexpected happens, but you don't need to understand the engine to be a good, careful driver. Parents don't need to become AI experts. They need a gradual release of the same core ideas we're teaching their kids, so home and school are speaking the same language.

Start by sharing your policies, plainly. Walk families through your school's actual guidelines, not just the philosophy behind them. At Exploris, we use a simple traffic light system with students: green means AI use is encouraged, for brainstorming, understanding concepts, or sparking creative ideas; yellow means use with permission, like drafting or checking grammar; red means never, which covers submitting AI work as your own, using it on tests, or hiding that you used it at all. When parents know this system, they can reinforce it at home instead of unintentionally undermining it, and they stop guessing at what's allowed.

Give them talking points, not just talking-tos. Most parents are unsure how to even start this conversation with their kids. Hand them a few concrete phrases and frameworks they can actually use:

• What AI is, in plain language. Artificial intelligence is software that learns patterns from huge amounts of information and then generates new content, text, images, answers, based on what it learned. It's worth naming the GIGO principle here too: garbage in, garbage out. What AI produces is only as good as what it was trained on and how it's asked.
• What AI can genuinely help with: brainstorming ideas, explaining a hard concept in a different way, generating a rough first draft to revise, or creating images for a project.
• What AI cannot do: it isn't always accurate and can confidently generate false information, it doesn't truly think or understand things the way a person does, it has no real experiences or emotions, and using it to do a student's actual work isn't a shortcut, it's cheating that gets in the way of real learning.
• A simple framework for evaluating anything AI produces. We teach our students the EVERY method: Examine what it generated, Verify the facts against reliable sources, Evaluate whether it actually answers the question, Refine the prompt if it doesn't, and remember You decide, because AI is a helper and the final judgment is always human. Parents can use this exact framework at the kitchen table.

Be honest about the real dangers, especially chatbots. This is the part that deserves the most care, and the part parents are most anxious about. AI chatbots can be genuinely helpful in small, low-stakes moments: I've shared with our families the story of a sixth grader who used AI to practice a nervous conversation with her soccer coach, building enough confidence to actually go ask him in person. That's AI doing what it should, helping a student rehearse, then stepping back so a human moment can happen.

But chatbots have real limits that every parent needs to hear plainly. Chatbots cannot recognize a genuine mental health crisis. They have no accountability and cannot provide real counseling. There have been documented cases of students confiding feelings of hopelessness to a chatbot that offered generic coping tips instead of alerting a parent, teacher, or counselor, because it simply cannot understand the context or urgency of a crisis the way a trained person can. Parents need to know that AI should never stand in for human connection when something serious is going on, and that the warning signs to watch for at home are the same ones we've always watched for, just showing up in a new place.

Give them a few simple habits to build at home, ones that don't require any technical expertise:

• Ask what tools your child's teachers actually approve for assignments, so home and classroom expectations line up.
• Ask your child to show you what they're learning about AI at school. Curiosity opens doors that lectures close.
• Keep repeating the core idea in your own words: AI is a helper, not a shortcut.
• Model the behavior yourself. Let your kids see you double-check an AI answer or ask a follow-up question for clarity.
• Talk openly about privacy, about what a chatbot can and can't understand, and about what responsible use looks like, the same way you'd talk about any other new privilege or tool.

There's a broader media literacy piece worth sharing too, because it protects kids well beyond AI. Every photo, video, and article, AI-generated or not, is a construct made by someone who chose what to include, what to leave out, and how to frame it. Helping our families understand that everything they see has a point of view, and that AI now makes it easier than ever to convincingly fabricate that point of view, builds the same muscle kids need to navigate AI responsibly.

When we bring families into this work instead of leaving them to figure it out alone, we're not adding one more task to an already full plate. We're making sure the values we're building in the classroom, curiosity over fear, guidance over policing, critical thinking over shortcuts, follow our students home. That consistency is what actually changes outcomes for kids

The Real Work Ahead

None of this is a finished checklist, it’s just a direction to start. Explore AI yourself first, so you can lead from experience instead of anxiety. Teach your staff to use it as a genuine workflow partner. Build guidelines rooted in trust and transparency instead of policing and detection. And help your teachers build AI literacy into the daily fabric of every subject, so our students grow into thoughtful, capable people who can think clearly with AI in the room, not people who've outsourced their thinking to it.

I've watched education survive calculators, spell check, CliffsNotes, the internet, and shared documents. Each time, we thought the sky was falling, and each time, we adapted and came out the other side with better teaching, not worse. I believe AI will be the same story, and I believe school leaders like us get to write how that story goes for our own students and staff. Let's change the game together, on purpose, instead of getting changed by it.

I'd love to hear how this is playing out in your own building. Reach me anytime at Deborah@Exploris.org.