Artificial intelligence (AI) is rapidly reshaping higher education. Its influence is increasingly evident in teaching, learning, assessment, research and institutional administration. Generative AI, in particular, has expanded the ability of students and staff to generate text, images, data analyses, presentations and computer code. Although these developments create important opportunities, they also raise concerns about academic integrity, data protection, algorithmic bias, unequal access and the continuing role of human judgement.
The central ethical challenge is therefore not whether higher education institutions should permit or prohibit AI. Rather, it is how they can adopt AI in ways that advance learning while protecting human agency, fairness and institutional accountability. UNESCO (2023) supports a human-centred approach in which AI enhances education without replacing the intellectual, professional and relational contributions of educators. For Regent Business School (RBS), responsible AI adoption requires more than regulating students’ use of generative AI. It requires an integrated institutional approach encompassing governance, AI literacy, assessment design, equitable access, data protection and meaningful human oversight.
Understanding AI in higher education
Artificial intelligence broadly refers to computer-based systems capable of performing tasks commonly associated with human intelligence, including language processing, pattern recognition, prediction, problem-solving and decision support. Generative AI is a category of AI that produces new content in response to user instructions or prompts. Its outputs can include written text, images, audio, presentations, data summaries and computer code.
In higher education, students can use generative AI to explain difficult concepts, generate ideas, improve the organisation of their writing and receive formative feedback. Academics can use it to develop learning activities, prepare teaching materials and create preliminary assessment questions. Administrative departments can employ AI-supported systems to respond to routine enquiries, analyse institutional data and identify students who could benefit from additional support.
These uses demonstrate that AI is not merely a mechanism for completing assignments. When used appropriately, it can become a learning and productivity tool. However, AI-generated outputs are based on patterns in existing data and can contain inaccurate, fabricated, biased or contextually inappropriate information. Consequently, the apparent confidence or fluency of an AI-generated response should never be mistaken for verified knowledge.
Opportunities for teaching and learning
AI presents several opportunities for RBS. It can provide students with personalised explanations, language assistance and formative feedback. This could be especially valuable in a diverse learning environment where students possess different levels of academic preparedness and study through different modes of delivery. AI can also support accessibility by helping students reorganise complex information, convert material into alternative formats and engage with learning content at their own pace.
For lecturers, AI can support the preparation of case studies, discussion questions, lesson plans and examples tailored to particular business contexts. It can also reduce the time spent on repetitive administrative activities, allowing academics to devote greater attention to student engagement, curriculum development and feedback.
These opportunities should not be viewed as automatic benefits. Their educational value depends on how the technology is selected and used. UNESCO (2024a) emphasises that educators require competencies in AI foundations, ethical use, AI-supported pedagogy and professional learning. RBS should therefore approach AI as a capability-development initiative rather than simply as a technological acquisition.
Ethical principles for responsible adoption
Responsible AI use in higher education should be guided by transparency, accountability, fairness, privacy, human autonomy and meaningful oversight. These principles must be translated into practical institutional requirements.
Transparency means that staff and students should know when AI has contributed to educational content, feedback or an institutional decision. Where AI is used to analyse student performance or identify students considered academically at risk, the institution should explain the purpose of the system, the information being processed and the limitations of its conclusions.
Accountability requires the institution to remain responsible for decisions supported by AI. Responsibility cannot be transferred to an algorithm or external technology provider. If an AI-supported recommendation affects admission, assessment, progression or student support, an appropriately authorised staff member should review the recommendation before action is taken.
Fairness requires institutions to examine whether an AI system disadvantages particular groups. AI tools can reproduce biases embedded in their training data or fail to interpret South African language patterns and educational contexts accurately. An automated system that appears neutral could consequently produce unequal outcomes for students from different linguistic, cultural or socioeconomic backgrounds.
Human autonomy means that AI should support rather than displace human judgement. Lecturers must retain responsibility for teaching, assessment and academic feedback. Students must also continue to develop their own ability to reason, analyse evidence, solve problems and communicate ideas. UNESCO (2021) argues that AI should be applied through a human-centred approach that protects human agency and advances inclusive education.
Meaningful human oversight further requires an opportunity to challenge AI-supported decisions. Students should be informed when AI materially influences a decision affecting them and should have access to an appropriate human review or appeal process.
Academic integrity and assessment
Generative AI has complicated established understandings of authorship, originality and academic integrity. However, AI use should not automatically be equated with plagiarism. Whether its use constitutes academic misconduct depends on the assessment instructions, the extent of its contribution, the accuracy of the student’s disclosure and whether the submitted work genuinely demonstrates the required learning.
For example, using AI to generate ideas or improve the structure of a draft could be permissible where the lecturer has authorised such assistance. Submitting substantially AI-generated content as one’s independent work would represent a different ethical issue. It could constitute unauthorised assistance, misrepresentation of authorship or failure to demonstrate the intended learning outcomes.
RBS therefore needs assessment-specific guidance rather than a single rule for every context. Each assessment should indicate whether AI use is prohibited, restricted or permitted. Where it is permitted, students should be told what uses are acceptable, how AI assistance must be disclosed and how generated information should be verified. The institution should also reconsider assessment practices by incorporating oral explanations, reflective accounts, applied projects, staged submissions and context-specific tasks that require students to demonstrate their reasoning.
AI-detection scores should not be treated as conclusive proof of misconduct. Such tools can produce unreliable results and should, at most, form one part of a fair academic-integrity investigation.
Privacy and data protection
The use of AI creates significant data-protection responsibilities. The Protection of Personal Information Act 4 of 2013 establishes requirements for the lawful processing of personal information and addresses the rights of individuals in relation to automated decision-making (Republic of South Africa, 2013).
Staff and students should not enter personal, confidential, commercially sensitive or institutionally protected information into unapproved public AI systems. This includes student records, research-participant information, unpublished research data, examination materials and confidential organisational documents. RBS should evaluate the data-processing practices of external AI providers before approving their institutional use. Agreements with service providers should clarify how data are collected, retained, transferred, secured and potentially used to improve AI models. Privacy should therefore be addressed at the point of technology selection and not only after a system has been implemented.
Equity and the South African context
AI adoption in South African higher education takes place within a system shaped by historical inequalities, uneven digital infrastructure and differences in students’ access to devices, connectivity and academic support. Woldegiorgis and Chiramba (2025) demonstrate that historically disadvantaged students continue to experience financial, linguistic and epistemic barriers within South African higher education.
Requiring students to use paid AI systems without providing institutional access could deepen existing inequalities. Students with premium tools, reliable connectivity and stronger digital competencies would possess advantages over students relying on restricted free versions or unstable internet access. RBS should therefore assess the accessibility and affordability of AI tools before embedding them in teaching and assessment.
Equity also requires attention to AI literacy. Providing access to technology is insufficient if students do not know how to question AI outputs, identify misinformation, protect personal information and disclose AI assistance. Institutional access should consequently be accompanied by structured training and ongoing academic support.
A responsible institutional approach for RBS
RBS should begin by establishing a reliable understanding of current AI use and awareness across the institution. A baseline survey or consultation involving students, academics, supervisors and professional staff could identify existing practices, training requirements, perceived benefits and areas of concern. This would prevent institutional policy from being based on untested assumptions.
The findings should inform an integrated institutional strategy containing six elements:
- Governance and accountability: Assign oversight to a multidisciplinary AI governance working group or an appropriate existing committee, with representation from academic leadership, teaching and learning, information technology, quality assurance, ethics, legal or compliance functions and the student body.
- Clear institutional guidance: Develop policies explaining acceptable, restricted and prohibited uses of AI in teaching, assessment, research and administration.
- AI literacy: Provide compulsory orientation and continuing development opportunities that teach staff and students how AI works, where it can fail and how it should be used responsibly.
- Responsible assessment: Require assessment instructions to communicate the permitted level of AI assistance and the relevant disclosure requirements.
- Data governance: Approve AI systems only after evaluating privacy, security, data ownership, retention practices and compliance with the Protection of Personal Information Act.
- Continuous evaluation: Monitor the educational effects of AI adoption, review unintended consequences and revise institutional practices as technologies and regulatory requirements develop.
This approach would move RBS beyond a narrow focus on misconduct and towards an institutional culture of responsible, critical and educationally meaningful AI use.
Conclusion
AI offers higher education significant opportunities to enhance learning, accessibility, academic support and institutional efficiency. Nevertheless, these opportunities are accompanied by risks involving inaccurate information, academic integrity, privacy, algorithmic bias, unequal access and excessive dependence on automated systems. The appropriate institutional response is neither unrestricted adoption nor blanket prohibition.
For RBS, responsible adoption should be grounded in human-centred education, transparent governance, fair access, data protection, AI literacy and meaningful human oversight. AI should strengthen the capacity of lecturers and students to teach, learn, reason and create; it should not replace these human responsibilities. By developing a coherent and evidence-informed institutional approach, RBS can position itself as a responsible contributor to the emerging AI landscape in South African higher education.
