You’ve Watched a Lot of Educational Technology Come and Go. Here’s How to Know AI Is Different.

If you have been a principal for any real length of time, you have a graveyard.

Not a literal one. But a professional one. The initiative that arrived with a district-wide rollout, consumed two years of professional development, and quietly disappeared when the next superintendent came in. The platform that was going to transform how teachers collaborated, used enthusiastically for a semester, then abandoned when the vendor stopped supporting it. The program that was supposed to close the achievement gap, the one before it, and the one before that.

Experienced principals have something most of the AI conversation does not account for: a track record to compare this to.

And that track record is the right starting point for evaluating what is actually happening with AI in schools.

Here is what I want to say directly: the healthy skepticism you earned through the last several waves of educational technology is an asset. Not an obstacle. The veteran principal who walks into an AI vendor presentation with a list of specific questions — not dismissal, not enthusiasm, specific questions — is protecting their school from another expensive failure.

But the same experience that earns that skepticism also has to reckon with something: AI is not the same category of thing as what came before it. And knowing the difference matters.

What the Previous Waves Had in Common

Most of the educational technology waves that did not stick shared specific characteristics.

They required significant behavioral change from people who were already overwhelmed. They produced outcomes that were difficult to attribute — hard to connect the platform adoption to anything specific that improved for students. They were dependent on vendor relationships, licensing renewals, and implementation support that districts could not sustain. And they were adopted at the district level for reasons that had more to do with optics and grant requirements than with the specific daily reality of what happens in buildings.

Experienced principals can run that checklist in their sleep.

Where AI Is Genuinely Different

AI does not require behavioral change in the same way. The most useful current AI tools work with what principals are already doing — drafting a communication, analyzing a data set, organizing meeting notes — and reduce the time those tasks take. There is no new workflow to adopt before value arrives.

The research on what AI is actually automating in knowledge work is also specific in a way previous technology waves were not. The tasks being automated are documented and measurable: compliance reporting, communication drafting, data analysis, scheduling optimization. The time return is real and trackable in a way that “improved collaboration” or “personalized learning” never was.

And AI is not tied to a single vendor relationship the way previous platforms were. The technology is foundational — built into the infrastructure of how information is processed — in a way that makes it more like the transition to email than the transition to any particular platform.

The Questions That Protect Your School

Experienced skepticism applied precisely is what saves schools from expensive failures. Here are the questions worth carrying into every AI conversation.

Does this tool address a task I am actually spending significant time on? Not a task the vendor assumes principals spend time on — a task you specifically are spending time on right now. If the answer is yes, the next question is whether it reduces that time or adds a new layer of management on top of it.

What does this tool do with student data? Specifically. Not in terms-of-service language — in plain sentences describing what information it accesses, where it goes, and who has it. If the vendor cannot answer that clearly and quickly, the conversation is over.

What does this tool make possible that I cannot currently do? Not what it replaces — what it enables. If the honest answer is “the same things I am currently doing, slightly faster,” that is useful information. If the answer is “things I have not been able to do because administrative tasks have been consuming the time,” that is a different conversation entirely.

The graveyard you have built over years of leading through failed initiatives is not a reason for dismissal. It is a reason for precision.

The principals who will use AI well are not the ones who adopt everything. They are the ones who evaluate specifically, adopt deliberately, and protect their schools from the next wave of tools that will arrive promising to transform everything.

That kind of precise judgment is built through experience.

Which means you are better positioned for it than you probably realize.