Educators must lead edtech choices, not vendors

educators must – American schools spent roughly $30 billion on educational technology in 2024, a figure projected to nearly double by 2033—yet teachers and students often say they’re not meaningfully consulted. In California State University’s OpenAI contract and Los Angeles U

In 2024, American schools spent roughly $30 billion on educational technology. By 2033, that number is projected to nearly double. For superintendents, it can feel like the calendar is already spoken for—emails, brochures, and demos from education technology companies arriving again and again.

But in the classroom, the questions don’t sound like product briefings. Teachers and students are trying to understand how to make sense of technology decisions that arrive from above—often before they’ve had a real chance to shape the problem those tools are supposed to solve.

The argument for a different process isn’t a rejection of technology. It’s a reordering of authority: who gets to name the gap. who decides what counts as a solution. and who holds the power when the school says yes or no. Schools already have rigorous pathways for tech adoption—pilots, data reviews, stakeholder check-ins, procurement policies. The trouble is that. by the time many schools run through those steps. they have already answered an earlier question: who gets to define what the school needs. That answer often comes from someone outside the classroom, sometimes before a teacher has even been asked.

Right now, the adoption path tends to start somewhere else. Someone—often a key decision-maker—is made aware of an educational technology solution through an email pitch. conference demo. or marketing campaign. The school then attends a demo. checks against its budget and procurement policies. may start with a pilot. and then adopts the tool.

In that sequence, the vendor is largely the center of the conversation. The question becomes: Do we adopt this tool?. The vendor explains why it should be adopted. and the school decides whether that explanation matches a need the school has already defined. Even when schools do evaluate what they buy. the search is constrained—often by where the vendor tells them to look. If a vendor pitches a tool for managing custodial schedules. for example. the school checks whether its custodial scheduling system is sufficient. not whether the broader educational problem is the one the tool will solve.

Federal and state edtech funding and grants can make the mismatch harder. Those funding streams narrow the scope of focus based on funding priorities rather than pedagogical ones. pushing schools toward whatever categories fit the grant language. And because decision-making is concentrated at the leadership level, vendors focus their selling on administrators rather than teachers. The buyer and the user are split structurally, especially when teacher voices aren’t driving the process.

That is how a “needs assessment” can end up running in reverse: the vendor names the gap, the school confirms it, and stakeholder consultation turns reactive instead of something schools build from the beginning.

Tensions aren’t hypothetical. They’ve surfaced across classrooms and institutions.

When the California State University system signed a contract with OpenAI, the announcement caught faculty and students off guard. More broadly. a 2025 survey found that 71 percent of faculty say administrators “overwhelmingly” lead conversations about introducing AI into research. teaching. and policy. with “little meaningful input” from faculty. staff. or students.

In K-12 schools, the pattern shows up in procurement too. More than 60 percent of teachers say they believe they should be the primary decision-makers regarding classroom technology. Yet only 38 percent report being consulted during the procurement process.

Los Angeles Unified’s experience with AI illustrates what can happen when the process moves fast without buy-in. Within three months of its high-profile launch of a $6 million AI chatbot in March 2024. the vendor collapsed and the tool was pulled. The teachers’ union then demanded that future AI tools be subject to transparent consultation with educators and to collective bargaining.

Against that backdrop, the proposed shift is to flip the entry point.

Instead of bringing a room full of brochures. demos. and swag. imagine a group of students and teachers sitting together—chosen as representatives of the population. not for enthusiasm about a technology. There would be no predetermined theme and no goal tied to a demo. The task would be straightforward: share what the biggest problems, struggles, and gaps in the school actually are.

The process matters because it changes how “needs” are defined. A teacher- and student-led needs inventory could become the first document in any edtech RFP. written before any external party is contacted. That document would name the gap the school is trying to close in the school’s own language. not a vendor’s.

When companies respond. they would have to contextualize their products against what the school actually wrote—rather than pitching against a generic value proposition. If the gap a school identifies is that students cannot sustain disagreement in classroom discussion. vendors would need to explain how their tool addresses that gap. or honestly acknowledge that it does not. The conversation would move from “here’s what we sell” to “here’s how we fit what you actually need.”.

The same group that writes the inventory would also sit on the evaluation panel. If teachers and students write the criteria up front and weigh in on which response best meets them, the buyer and the user stop being structurally separated.

That approach would be slower at first. But slowness being a negative assumes the pace now—driven by vendor frameworks and marketing cycles—is worth defending. and there is “no evidence” to support that speed. The case being made is that current velocity is often the consequence of steps being skipped. not proof that the process is more efficient.

There is also a risk in expecting every problem to be fixable with software. Some answers won’t translate cleanly into a product. The honest outcome might be “we need smaller classes” or “we need an additional counselor.” Finding that early is cheaper than rolling out a failed platform that solves the wrong problem.

Underlying all of this is a concern about cognitive outsourcing. Using vendor-provided frameworks can feel comfortable and fast, but it can subtly erode school culture over time. The proposed fix is not just procedural. It goes to agency—who schools treat as reliable authorities in classrooms. and who gets to set the terms before a vendor enters the room.

The story here isn’t that technology has no place in schools. It’s that classrooms shouldn’t have to adapt to technology choices made elsewhere. If the people closest to the day-to-day realities of teaching and learning define the questions first. schools may still end up choosing platforms. They may still change schedules. But they won’t be building solutions around someone else’s pitch.

educational technology edtech procurement teacher voice student input AI in education OpenAI contract Los Angeles Unified AI chatbot California State University RFP educators agency