Open Sourcing Trust: What It Takes to Build Public Infrastructure for AI-Ready Higher Education

As AI reshapes higher education, the biggest opportunity may not be the models themselves – but the shared infrastructure that makes them useful, trustworthy, and sustainable.

AI is rapidly moving from experimentation to institutional strategy.

Across higher education, colleges are exploring how AI can strengthen advising, streamline operations, personalize student support, and help staff make better decisions. The promise is significant – but so is the realization that AI is only as effective as the systems beneath it.

Over the past year, we’ve seen the same conversation emerge across very different communities. From open-source developers at FOSDEM, to practitioners at Good Tech Summit, to global leaders in digital public infrastructure at ICT4D and UN Open Source Week, one message has become increasingly clear: before organizations can fully benefit from AI, they need stronger foundations. Better data. Shared standards. Trustworthy governance. Infrastructure designed to connect – not further fragment – the technologies institutions rely on every day.

Whether the conversation begins with open source, digital public infrastructure, or AI, it increasingly arrives at the same conclusion: technology advances faster than the institutions, governance, and shared infrastructure required to sustain it.

That perspective is increasingly shared across philanthropy as well. In a recent reflection on the future of postsecondary success, Gates Foundation’s Patrick Methvin argues that AI should help institutions surface insights, improve operations, and free up staff to focus on students – but that the goal isn’t simply more technology. It’s building systems that work better for students and the people who support them. The same principle applies here. AI in higher education will ultimately depend on the strength of the infrastructure beneath it.

For colleges and universities, this isn’t simply an AI challenge. It’s an infrastructure challenge.

From AI Excitement to Infrastructure Reality

Most colleges and universities don’t lack technology – they lack technologies that work together. A student’s admissions record may live in one system. Financial aid information in another. Advising notes somewhere else. Learning management data in a fourth platform. Career services, student success initiatives, and workforce outcomes often exist in separate applications owned by different offices.

When information remains fragmented, students may have to repeat their stories across offices or piece together their own experiences across disconnected systems, while advisors spend valuable time searching for the information they need instead of supporting students. Institutional researchers build workarounds to connect data across systems. Institutional leaders struggle to see the full picture. And AI systems inherit the same disconnected data that humans already struggle to navigate. Adding another AI tool doesn’t solve fragmentation. It can automate and scale the disconnect, making the gaps in underlying infrastructure more systematic rather than solving them.

That’s why conversations are increasingly shifting beyond individual applications toward data models, standards, governance, and interoperable systems that create trusted foundations for innovation. Infrastructure isn’t the technology people notice first. It’s the technology that determines whether everything else works – and when those foundations are built to be open, interoperable, and broadly shared, they become public infrastructure that benefits the entire ecosystem.

Open Source Is Only the Beginning

Open source represents something bigger than freely available code or data. At its best, it enables transparency, collaboration, replicability, adaptability, and shared ownership. It allows communities facing similar challenges to solve problems together, creating knowledge and solutions that others can build on rather than reinventing them independently.

We’ve seen this firsthand through our participation in open-source and digital public infrastructure communities – from FOSDEM to Good Tech Summit, ICT4D, and UN Open Source Week – where one lesson has consistently emerged: successful open-source initiatives aren’t sustained by code alone. They depend on governance, documentation, community participation, and long-term stewardship.

Higher education has long embraced these same collaborative values. But experience has also shown that openness alone doesn’t guarantee impact. Open-source projects often struggle with sustainability, implementation support, governance, and adoption. Valuable technology can remain underutilized if institutions lack the resources – or the community – to successfully implement and maintain it.

The question isn’t simply whether software is open. It’s whether it becomes infrastructure the community can depend on.

What We’ve Learned About Building Public Goods

Over nearly a decade of building open data infrastructure and digital public goods across higher education, global health, climate, economic opportunity, and open research, we’ve learned that technology is rarely the limiting factor. The harder challenge is building the institutions, governance, and communities that allow shared infrastructure to endure.

Across every initiative, one lesson has remained consistent: open source creates possibility. Shared standards, community participation, and long-term stewardship create sustainability.

Public infrastructure succeeds not because every organization uses the same technology, but because they share common foundations to work together.

Three lessons continue to shape our approach:

Build with institutions – not for them. The best infrastructure reflects the realities of the people who use it every day. That means involving institutions throughout the design process, testing ideas early, incorporating feedback from the communities expected to depend on it, and recognizing that technical decisions often have operational consequences. This kind of co-design doesn’t slow innovation. It makes adoption possible. It was one of the strongest themes we heard at Good Tech Summit, ICT4D, and UN Open Source Week, where leaders across philanthropy, government, and civil society emphasized that trustworthy infrastructure isn’t built in isolation – it emerges through continuous collaboration with the communities expected to rely on it.
Shared standards matter more than any single application. Institutions don’t all need the same technology. They need shared meaning across the technologies they already use. Common data structures, interoperability standards, and thoughtful governance allow different systems to exchange information consistently and securely. They don’t require institutions to use the same software. Instead, they allow different technologies to interpret and exchange information consistently. Healthcare’s adoption of standards like Fast Healthcare Interoperability Resources (FHIR) demonstrates how shared foundations can unlock innovation across an entire ecosystem. Higher education has a similar opportunity to strengthen how information flows across institutional units – such as advising, academics, financial aid, workforce pathways, and student support – not by requiring every institution to use the same tools, but by enabling different tools to work together. Trust begins long before an AI model generates an answer. It begins with consistent, connected data – and with the shared standards and governance that make that data reliable.
Communities sustain infrastructure. Making technology open is only the beginning. What turns it into public infrastructure is the community that governs, maintains, and improves it over time. Governance determines who participates. Stewardship determines whether it evolves. Long-term investment determines whether it survives. The most durable public goods are continuously maintained, governed, and improved by the people and institutions that depend on them.

Broad Access Is Infrastructure

Which institutions have the capacity to adopt new technologies? Can smaller or resource-constrained colleges access improved systems without significant new investments? Are systems designed to work across different institutional contexts and the populations they serve – not just the most well-resourced ones serving traditional students?

These aren’t questions to answer after deployment. They’re decisions that shape infrastructure from the very beginning. Building for broad participation ultimately creates stronger technologies – for everyone.

Infrastructure is never neutral. Every decision about governance, standards, participation, and stewardship shapes who can participate, whose needs are reflected, whose voices influence what comes next.

A Broader Movement Toward Digital Public Infrastructure

These conversations now extend well beyond higher education.

Across global development, healthcare, philanthropy, and public services, organizations are reaching remarkably similar conclusions:

Responsible AI depends on trustworthy data.
Trustworthy data depends on interoperable infrastructure.
Interoperable infrastructure depends on shared standards, governance, and institutions willing to steward those foundations over time.

Those conversations are increasingly converging around digital public infrastructure – not simply as a collection of technologies, but as the institutional foundations that allow innovation to scale responsibly.

Why This Matters Now

AI capabilities are advancing faster than institutional infrastructure, while institutions face growing pressure to figure out how to adopt AI. Without the necessary foundation, even promising AI applications can be ineffective – or reinforce the very gaps they are meant to address. That gap presents both a challenge and an opportunity.

Without stronger foundations, institutions risk reinforcing existing silos, increasing implementation costs, and limiting AI’s ability to meaningfully improve student outcomes.

But by investing in shared infrastructure today, higher education has the opportunity to build technologies that are more interoperable, more accessible, more sustainable – and ultimately more human-centered.

Building the Foundation Together

The future of AI-ready education won’t be built by any single institution or organization. It will emerge through collaboration around shared standards, open infrastructure, and digital public goods that strengthen the broader ecosystem.

AI may be today’s catalyst. But the lasting opportunity is much bigger than AI itself.

It’s creating the public infrastructure that allows innovation – and student success – to scale for years to come.

This article is the first in a broader conversation about what it takes to build AI-ready education systems. In the months ahead, we’ll explore the role of interoperable infrastructure, governance, institutional readiness, and open collaboration in helping higher education move from AI experimentation to lasting impact.

We hope these ideas contribute to a broader conversation already taking shape across higher education, philanthropy, open source, and digital public infrastructure communities. If you’re exploring these challenges within your own institutions or organizations, we’d welcome the conversation at education@datakind.org.

Image above courtesy of iStock.

Join the DataKind movement.

Interested in supporting our work? Donate here.
Interested in sponsoring a project? Partner with us.
Interested in subscribing to our newsletter? Sign up.