How higher education can embrace AI while protecting academic integrity
The rise of artificial intelligence (AI) has significantly changed the education landscape across higher education institutions in South Africa. It has also caught a number of institutions off guard, with some still considering how best to tap into AI while managing the challenges and potential risks it presents. Many institutions have begun drafting policies and frameworks to guide how they can successfully integrate AI into their curricula.
Exploring ways of embracing AI
Fundi, in partnership with Mail & Guardian, recently hosted a conversation to explore practical ways in which higher education can respond to AI-related challenges. One of the critical questions was how universities can adopt and harness the power of AI without compromising the integrity and credibility of their academic programmes.
Fundi’s chief executive officer, Mala Suriah, welcomed participants and outlined the objectives of the event. Held at Thaba Eco Hotel, south of Johannesburg, the summit was attended by the University of Cape Town’s vice-chancellor, Professor Mosa Moshabela; Dr Linda Meyer, managing director of IIE Rosebank College; and Professor Suzanne Graham, professor of international relations at the University of Johannesburg.
Fundi’s chief executive officer, Mala Suriah
Creating a platform for higher education leadership
Suriah said the conversation afforded higher education leadership an opportunity to share ideas on how AI can be woven into the fibre of the sector’s ecosystem. She said Fundi supports the view that technology, particularly AI, has the potential to help address access to higher education.
“We hope the responsible inclusion of AI will assist our business to achieve its purpose,” said Suriah. “We have put together this platform to collectively look at policy direction, regulations and frameworks so that we can pursue a clear direction on AI.”
Professor Brian Armstrong
Embedding AI into the curriculum
Professor Brian Armstrong, a respected ICT expert, academic and non-executive director specialising in digital business strategy, set the tone for the discussions with a keynote address. He later moderated a panel discussion featuring the three academics, who fielded questions on how their institutions are tapping into the benefits of AI.
Professor Armstrong traced the evolution of AI and its impact on humanity. He said AI is already part of daily life and is going to have a direct and profound impact on personal, professional and social lives.
He cautioned that technology should never be an end in itself, but rather a means to an end. Professor Armstrong said the critical challenge facing the higher education sector is how institutions can leverage AI to innovate and enhance teaching and learning.
“We need to be teaching AI across all faculties and integrating it into normal curriculum processes,” he advised. “It should be added into other disciplines such as medicine, law and engineering, to mention a few.”
AI may outperform humans
AI can be used for better or for worse, said Professor Armstrong, adding that the extent to which it is used positively or negatively depends on people. He said AI is developing at a rapid pace and that universities should plan for where it is going, not only where it is now.
He shared some of the breakthroughs in AI, saying that while humans are still ahead in key areas such as emotional intelligence, indications are that AI may surpass human intelligence in certain areas in the near future. He cited the advanced capabilities of some of the latest models, including Claude, Gemini, DeepSeek and ChatGPT, as examples of how quickly AI is progressing.
Influencing the content and direction of AI
Professor Moshabela supported the view that the higher education sector should adopt AI as part of charting a path towards a digital future. He said the Global South should actively participate in influencing the content, coding and direction of AI so that “our African values are reflected in it”.
He said Africa should build its computing capabilities in order to leapfrog countries that are currently leading in AI. Moshabela said UCT has existed for about 200 years and has been able to navigate the challenges it has faced by adapting and pivoting quickly.
He added that the university has ensured that its curriculum remains responsive and relevant to social needs. Universities, he said, should redesign their curricula and embed AI so that they can produce graduates who are future-ready.
Protecting intellectual capacity
For Professor Meyer, the adoption of AI by universities is beyond dispute. The issue is how to navigate questions of trust and ethics while ensuring innovation and integrity.
Universities should carefully consider how students can use AI without outsourcing their agency and thinking, Professor Meyer argued. Students, she added, should learn and sharpen their cognitive skills without defaulting to AI.
“We should avoid a situation where we produce graduates who are proficient in AI and yet cognitively shallow. They should not surrender their intellectual capacity to it,” she cautioned.
Advancing the future of graduates
“We need to be pragmatic and think about what will be required from our graduates,” said Professor Meyer. She warned the higher education sector against burying its head in the sand and pretending AI is not here.
“We need to understand what the challenges are and ensure we have sufficient guardrails in place so that we can work with AI and avoid chaos. We need to embrace what will advance the future of our students,” said Professor Meyer.
She said that if employers are not satisfied with the quality or level of skills that universities are producing, graduates will find it difficult to secure employment. “So, as universities, we need to be thinking about assessment methodologies differently to make sure they are fit for purpose,” she added.
Developing an academic voice
At UJ, said Professor Graham, students are encouraged to use different technologies such as Copilot and Grammarly. These tools assist with language construction and editing, particularly for students who speak English as a second language.
She said UJ recommends that students use AI for initial literature scans, adding that the tools also serve to create awareness. “But what we don’t want our students to do is use it to write for them, as this takes away their independent thought and originality. We want our students to be critical thinkers who engage with literature and develop their own academic voices,” she said.
Discouraging a culture of intellectual dishonesty
Professor Graham said AI should be used across various disciplines, including STEM subjects and the humanities. For instance, students can use AI to design architecture in rural spaces or develop sustainable tools for climate change, she suggested.
She said one of the critical questions UJ is currently grappling with is how to teach students to use AI ethically and with integrity. Some students use AI in ways that amount to plagiarism, even though many do so unintentionally, she said.
Through the use of various AI tools, lecturers are able to detect instances of cheating by students, she said. Instead of simply penalising them, Professor Graham said, the university teaches students that plagiarism is not allowed in the academic world.
“We teach them to reference their work and develop a culture of promoting and upholding academic honesty and integrity,” concluded Professor Graham.