As generative AI continues to reshape the global workforce, South African universities are moving beyond mere “guidelines” to establish world-leading ethical frameworks. Rather than banning tools like ChatGPT, local institutions are adopting a human-centred approach that prioritizes AI literacy over surveillance.
Leading the charge is North-West University (NWU), which has officially become the first in the country to transition from temporary guidelines to a formal AI Framework Policy.
Three Pillars of the South African Approach
1. Human-Centred Literacy (NWU)
Professor Anné Verhoef, director of the NWU AI Hub, emphasizes that technology should serve society, not the other way around.
- Student Support: NWU launched a free online course, “AI for Academic and Career Success,” to teach students about algorithmic bias and Western data dominance.
- Staff Empowerment: Lecturers are trained through the “Winning the AI Assessment Game” course, helping them decide when to embrace AI and when to stick to traditional methods.
2. Ethical Choice over “Policing” (Stellenbosch)
Stellenbosch University (SU) has taken a bold stance by discontinuing Turnitin’s AI detection software. Dr. Hanelie Adendorff argues that automated detection is often unreliable and lacks pedagogical value.
- The “Thinking Tool”: SU uses an “ethical position statement” that focuses on authenticity and transparency.
- Student Responsibility: Students must sign declarations and maintain an ongoing dialogue with supervisors, shifting the focus from the final product to the process of learning.
3. Innovation over Intimidation (UKZN)
At the University of KwaZulu-Natal (UKZN), the philosophy is built on trust. Professors Nyna Amin, Donrich Thaldar, and Thabo Msibi advocate for “education instead of enforcement.”
- Trust First: UKZN views AI as a transformative opportunity, encouraging students to use it as they would any other research tool—without the fear of constant surveillance.
Defining the Line: Editing vs. Ghostwriting
The consensus across South African academia is clear:
- Permissible: Using AI to enhance English proficiency or brainstorm structures (provided it is declared).
- Unethical: “Ghostwriting,” where AI produces the entire work and the student claims it as their own.
By fostering “AI-literate” graduates, South African universities are ensuring that the next generation can navigate a tech-driven world without sacrificing the integrity of human thought.
