Built to Evolve

Schools are buying more technology.

The more interesting question is what they are buying it for.

For years, education technology procurement largely revolved around functionality. Could a system manage admissions? Deliver content? Track attendance? Assess learning? Connect teachers and students?

Those questions still matter.

But artificial intelligence is adding another.

What are we buying today that can still evolve with us tomorrow?

It is a subtly different test.

Technology cycles have shortened. Capabilities that once required specialist software increasingly appear within existing platforms. AI models improve, interfaces change and new applications emerge faster than the traditional institutional planning cycle.

That does not make long-term technology decisions obsolete.

It changes what “long term” needs to mean.

The numbers illustrate why.

Gallup reported this year that six in ten US public-school teachers use AI for their work. Yet only 18 per cent said they had received formal guidance from school administrators on its use.

In higher education, the Digital Education Council surveyed 45,398 students and faculty across 35 countries. Among students, 43 per cent said AI had not been integrated into any of their courses. Of those who had experienced integration, 5 per cent said it had transformed how they learn.

Only 29 per cent thought their instructors were well equipped to guide them on AI use.

These figures do not necessarily point to slow adoption.

They suggest something more interesting: technology adoption and institutional integration are not the same thing.

One can happen remarkably quickly. The other involves systems, people, governance, training and choices about how technology contributes to education itself.

That distinction is beginning to reshape the market.

The National University of Singapore offered one example this week. Its new collaboration with OpenAI will make ChatGPT Edu available across its community of students, faculty and staff. But the agreement extends beyond access to a platform. It encompasses education, research, administration and AI talent development.

The distinction matters.

Are institutions buying products, platforms or capabilities?

Increasingly, the answer may be some combination of all three.

A product solves a defined problem. A platform connects multiple activities. Capability determines whether an institution can continue extracting value as both evolve.

For education technology companies, this creates an equally interesting shift.

As sophisticated technology becomes more widely available, differentiation is unlikely to rest on access alone.

Implementation matters. Integration matters. Evidence matters. So do trust, expertise and an understanding of the institution in which the technology will operate.

The relationship between buyer and provider consequently begins to look different.

From vendor to partner. From licence to capability. From procurement to transformation.

This is not unique to education.

Enterprise technology has been moving in this direction for years. Cloud computing shifted organisations from owning infrastructure towards consuming services. Software-as-a-service changed how businesses bought and updated applications. AI is accelerating another iteration of that transition.

Education, however, introduces a particular set of considerations.

Decisions about technology sit alongside questions of learning outcomes, assessment, safeguarding, personal data and academic integrity.

The calculation is therefore not simply whether a technology works, but where it creates enough value to justify changing the way an institution works.

That makes the appropriate model less obvious.

Some institutions will build capabilities internally. Others will rely more heavily on specialist partners. Many will combine large technology platforms with education-specific providers and their own institutional expertise.

The strategic question is not which model is universally right.

It is which capabilities an institution needs to own, and which are better accessed through partnership.

That question may prove more durable than today’s debate about individual AI tools.

It also changes the criteria by which technology partnerships are judged.

Functionality and price remain important. But adaptability, interoperability, implementation capability and evidence of outcomes increasingly belong in the same conversation.

Because the useful life of a technology partnership may no longer depend simply on how long the product remains unchanged.

Its value may depend on how successfully it can change.

Perhaps that is the more useful way to think about longevity in technology.

The question is not whether what institutions buy today will look the same in five years. It almost certainly will not. The question is whether it will still be valuable when it doesn’t.