The debate around student funding in South Africa has entered a new phase.
In recent weeks, calls to scrap NSFAS have grown louder. Political leaders, education commentators and policy analysts have argued that the institution has become synonymous with administrative failures, governance crises and recurring uncertainty for students. The argument is no longer simply about reforming NSFAS; it is increasingly about replacing it altogether.
Given NSFAS’s recent placement under administration for the third time in less than a decade, these frustrations are understandable. Questions about governance, payment delays and institutional stability continue to dominate headlines. But amid this debate, I believe we are overlooking a more fundamental question.
Even if NSFAS is scrapped tomorrow, what replaces it?
Whether student funding is administered by NSFAS, universities themselves, a decentralised model, or an entirely new institution, South Africa will still require a system capable of processing hundreds of thousands of applications, verifying eligibility, identifying fraud, distributing billions of rands and responding to student appeals. In other words, changing the institution does not automatically solve the administrative challenge.
The challenge remains.
And that is precisely where Artificial Intelligence deserves serious consideration.
In 2024, I conducted research among University of Johannesburg Humanities students to understand their perceptions of integrating AI into NSFAS fund distribution. Contrary to what some might expect, students were not overwhelmingly hostile to the idea. They expressed concerns about errors, corruption and the possibility that AI could overlook unique student circumstances. Yet these concerns did not translate into outright rejection. Students across different socio-economic backgrounds, school types and levels of technological familiarity displayed remarkably similar attitudes.
The most important finding from my research was that students were not rejecting AI. They were questioning whether any system could be trusted to work fairly.
That distinction matters.
Too often, discussions about AI assume public resistance to the technology itself. My findings suggest something different. Students appear willing to embrace technological innovation if it improves efficiency, transparency and accountability. Their scepticism is directed less at artificial intelligence than at institutional failure.
The years since my research have only reinforced that conclusion.
South Africa’s broader conversation around AI has matured considerably. Government has begun exploring national AI governance frameworks, major technology firms are investing hundreds of millions of rand into AI infrastructure and skills development, and public institutions across the world are increasingly experimenting with AI-assisted administration. At the same time, the lessons of AI failures have become equally clear. South Africa recently withdrew a draft AI policy after concerns emerged regarding AI-generated references, demonstrating that technology without human oversight can undermine public trust rather than strengthen it.
This is precisely why the debate should not be framed as a choice between humans and machines. The future is neither fully automated nor entirely manual. It is hybrid.
The most promising evidence emerging from public-sector AI implementation internationally suggests that AI works best when it augments human decision-making rather than replacing it. Properly governed systems can reduce processing times, improve administrative efficiency and assist officials in managing large-scale workloads while retaining human accountability.
Student funding administration is exactly the kind of environment where such an approach could prove valuable. Imagine a system where AI assists with application verification, flags suspicious claims, identifies duplicate submissions, predicts payment bottlenecks and provides real-time communication to students. Human officials would remain responsible for final decisions, appeals and exceptional cases, but their work would be supported by tools capable of processing information at a scale no bureaucracy can achieve alone.
This is not a fantasy.
It is increasingly becoming standard administrative practice across governments and institutions worldwide. Critics will rightly point out that AI can reproduce bias, make mistakes and create new forms of exclusion.
They are correct.
But so can human systems.
The question is not whether AI is flawless. It is not.
The question is whether we can design governance frameworks that make AI more transparent, auditable and accountable than many of the administrative processes students currently experience.
My research suggests that students are open to that possibility.
That finding should not be ignored.
As South Africa debates the future of student funding, policymakers must resist treating institutional reform and technological innovation as separate conversations.
They are deeply interconnected.
If NSFAS is retained, it will require more effective administrative systems.
If NSFAS is scrapped, its replacement will require more effective administrative systems.
Either way, the challenge remains the same.
The debate, therefore, should not be whether AI belongs in student funding administration.
The debate should be how to ensure that any future system, whether NSFAS or its successor, uses technology responsibly, transparently and in ways that genuinely serve students.
My research found that students are ready for that conversation.
The question is whether policymakers are.
Sbusiso Gwala is an Assistant Lecturer at the University of Johannesburg, a Cum Laude Master’s graduate, youth leader, and writer on education, faith, leadership, and social transformation.
