South African universities face a defining moment: they can police artificial intelligence on campus or teach students how to use it wisely. The choice will shape whether graduates enter the workplace prepared for the reality they'll face.
Universities across South Africa are rightfully tightening controls around unethical AI use in assignments. The concern is real. Students now have access to tools that can generate text, answer questions, solve problems and imitate original work at unprecedented speed. A qualification must still reflect a student's own competence, judgment and effort. But if institutional response stops at detection, surveillance and punishment, the sector will miss a more urgent opportunity.
The real distinction that matters is whether students understand the difference between using AI as a support tool and using it as a substitute for thinking. This distinction is not academic—it is economic. The workplaces these students are entering are not AI-free. Artificial intelligence is already shaping how people write, research, analyse, communicate and solve problems. Employers are not seeking graduates who have been kept away from AI. They want people who can use technology responsibly, question outputs critically, protect confidentiality, verify facts and still take ownership of the final result.
That is why the conversation must move from enforcement toward literacy. Students need clear rules, certainly, but they also need clear teaching. They must know when AI use is acceptable, when it is not, how it should be disclosed, and why ethical use matters. They must learn that convenience cannot replace competence. If a student cannot explain the logic behind an answer, defend a recommendation, or apply knowledge in a real setting, then the tool has not educated them—it has only assisted them to submit something.
The answer is not a simplistic choice between banning AI and embracing it uncritically. International guidance has already moved toward a human-centred approach to generative AI in education, with South African higher education research pointing to both the benefits of AI for teaching and learning and the urgent need for policy frameworks that protect academic integrity.
For institutions, this means rethinking assessment entirely. More oral assessments, workplace simulations, practical demonstrations, staged assignments, reflective submissions and supervised evaluations can all distinguish authentic learning from automated output. It also means making disclosure normal. Every assessment should state whether AI is prohibited, permitted in limited ways, or required for a defined purpose. Students should not be left to guess where the line is.
For students, the message must be equally clear: AI can help you learn, but it cannot learn on your behalf. It can assist with structure, brainstorming, summarising and refinement, but it cannot replace your responsibility to understand the work, test the information and stand behind what you submit.
This conversation carries wider significance for South Africa. The country faces pressure to build a more employable, adaptable workforce in a changing economy. That requires education providers to produce graduates who are digitally capable, ethically grounded and ready to work with new technologies without becoming dependent on them. The future will not reward those who avoid AI altogether. It will reward those who know how to use it wisely.
The current clampdown is a necessary first step, not the final one. But the long-term goal must be to build a culture in which students understand that technology is a tool, not a shortcut to credibility. Education still has to do what it has always done at its best: develop independent thinkers, capable practitioners and responsible citizens. AI changes the context, but it does not change that mission.