Frequently Asked Questions​

Stellenbosch University approaches AI as a teaching, learning and assessment issue rather than a technology or compliance issue. The focus is on supporting meaningful learning, academic integrity, responsible use and context-sensitive decision-making. AI use is considered in relation to the intended learning outcomes and the evidence required to demonstrate learning. 

There is no universal "yes" or "no" answer. Whether AI may be used depends on the purpose of the task, the learning outcomes involved and the conditions set by the lecturer. In some contexts AI use may be prohibited, restricted, allowed, encouraged or required. 

Regardless of whether AI is used, students remain responsible for meeting the required learning outcomes and academic standards and may be asked to explain or defend their work

Not currently. SU has an ethical position statement and guidelines that support responsible, context-sensitive decision-making. The approach is deliberately principle-driven rather than a fixed rulebook, because appropriate AI use differs across contexts, disciplines and assessment purposes.

The current guidelines are structured around four interrelated considerations: authenticity, fairness, accountability and transparency. These are intended as thinking tools rather than compliance checks. 

Lecturers should clearly explain whether, when and how AI may be used in a module or assessment. This should include what forms of AI assistance are allowed or not allowed, whether AI use must be declared, what students remain responsible for, and why particular conditions apply. The aim is to help students understand the learning purpose of the task, not only the rule. 

SU is developing an AI Use Bar to help academics make decisions about AI use. These decisions are informed by both pedagogical considerations (what students need to learn) and practical considerations (what can realistically be managed and assessed). The aim is to support learning rather than simply regulate tool use. 

No. SU discontinued the use of Turnitin's AI text detection functionality at the end of 2025 because of concerns about the reliability and validity of AI detection tools in high-stakes academic contexts. 

The focus is on assessment design, transparency and student accountability. Academics are encouraged to gather multiple sources of evidence of learning, which may include drafts, process evidence, oral explanations, in-class activities, reflections and other opportunities for students to demonstrate their understanding. 

Many academics are redesigning assessments to make student thinking more visible. This may involve requiring students to explain their reasoning, justify decisions, critique AI-generated outputs, apply knowledge in context, reflect on their process or defend their work orally. The goal is to strengthen the validity of assessment evidence. 

Students should follow the declaration requirements set out in the module, assessment instructions or postgraduate process. The purpose of declaration is to support transparency, accountability and responsible learning. 

A useful declaration does not always need to be a detailed list of tools or prompts. Depending on the task, students may be asked to explain how AI supported their work, how they checked the accuracy of AI-assisted outputs, what decisions they made themselves, and why the submitted work can still be regarded as their own. The central point is that students remain responsible for the accuracy, integrity and quality of the work they submit. 

No. Take-home essays and similar tasks can still play an important role, especially as opportunities for students to develop ideas, practise writing, receive feedback and deepen their understanding. However, they should now be designed with the assumption that they are AI-enabled contexts, because AI use cannot be fully controlled outside invigilated conditions. 

The key question is therefore not whether take-home tasks should disappear, but what purpose they serve in the broader assessment strategy. If they are used for formative learning, AI use may sometimes be appropriate and even useful. If they carry high-stakes summative weight, lecturers need to consider how the task will provide valid evidence of students’ own knowledge, skills and judgement. 

It does not simply mean replacing one assessment format with another. It refers to assessment strategy and design: making learning visible, helping students understand the consequences of outsourcing learning, and ensuring that high-stakes assessments still provide valid evidence of independent understanding and capability. 

Start by talking to the student. A conversation can help clarify what happened, how the student approached the task, and whether they can explain or defend the submitted work. 

Do not rely on a single source of evidence, and do not treat AI detection tools as sole or conclusive evidence. Consider multiple sources, such as the submitted work, drafts, process evidence, in-class work, oral explanation, or the student’s ability to account for their decisions. 

If, after engagement with the student, there is sufficient evidence of a possible breach of academic integrity, the matter should be handled through the normal academic integrity procedures. 

Students should declare AI use where this is required by the lecturer, module guidelines, assessment instructions or postgraduate processes. A declaration should indicate how AI was used and should support transparency and reflection. Declaring AI use does not remove the student’s responsibility for the accuracy, integrity and quality of the submitted work.

Yes. AI use may be prohibited where students need to demonstrate knowledge, skills or judgement without AI assistance. However, prohibition should be educationally defensible, clearly communicated, and practically enforceable. 

The key question is whether AI use would undermine the intended learning outcome or the validity of the assessment evidence. Where AI use is prohibited, lecturers should design the assessment conditions in ways that make this possible to uphold. 

Formative assessment supports learning as it develops. In these tasks, AI use may be appropriate if it helps students practise, receive feedback, test their understanding or develop AI literacy. It may also undermine learning, if relied on uncritically.  

High-stakes assessment, however, must provide credible evidence of what students know and can do. SU’s approach therefore asks lecturers to design assessment strategies where students can learn with support, but must also demonstrate the required knowledge, skills and judgement in valid and accountable ways. 

AI literacy involves understanding how to use AI responsibly and effectively. This includes evaluating AI outputs, recognising limitations and bias, verifying information, understanding ethical implications and knowing when AI use may support or undermine learning. AI literacy is increasingly viewed as an important graduate capability. 

Support includes institutional guidelines, workshops, faculty discussions, short courses, AI literacy resources and practical tools to help academics make informed decisions about teaching, learning and assessment in an AI-enabled world. 

No. Different disciplines use and produce knowledge in different ways. Appropriate AI use in a laboratory report, design project, literature review, programming task or professional placement may differ significantly. While institutional principles remain consistent, implementation needs to be context-sensitive. 

Several AI-related pilots and initiatives are under way across the university, including AI tutoring systems and other forms of AI-supported learning. These initiatives are being explored carefully with attention to educational value, data protection, fairness, transparency and accountability. 

AI can support learning through explanations, feedback, language assistance, study guidance and personalised support. However, students may also become over-reliant on AI or outsource the cognitive work required for learning. The aim is to help students use AI critically while continuing to develop disciplinary knowledge, judgement and independent thinking. 

There is not currently a single mandatory AI literacy certification for all academic staff. Staff have access to support opportunities, including two Digital Education Council courses, the ALAI short course, workshops, faculty-level discussions and other institutional resources. 

The emphasis is on contextual and disciplinary AI literacy rather than uniform tool knowledge. 

Postgraduate students are expected to discuss intended AI use with their supervisors, declare how AI tools are used, and take full responsibility for the accuracy, originality and integrity of the final work. Ethical and transparent AI use, aligned with supervisory guidance and disciplinary norms, is not in itself grounds for penalty. 

AI can both support and undermine equity. While it may improve access to explanations, feedback and multilingual support, AI systems can also reproduce existing biases and under-represent particular languages, regions and knowledge traditions. Responsible AI use therefore requires critical engagement with questions of representation, access, knowledge and power. 

One of the central challenges is ensuring that submitted work remains a valid representation of student learning. AI is prompting universities to reconsider what counts as evidence of learning, how learning is assessed and how students develop the knowledge, skills and judgement expected of graduates. 

SU does not currently operate large-scale, institution-owned generative AI infrastructure that would allow direct measurement of AI-related water or energy use. Most AI use happens through externally hosted third-party platforms, where institutional visibility is limited. 

AI prompts universities to reconsider what they are trying to cultivate in students. If higher education is reduced to producing polished outputs or credentials, AI can easily become a shortcut. If the purpose is to develop people who can think, judge, inquire, create, act ethically and contribute responsibly, then the learning process itself remains central. AI should therefore be considered in relation to the educational purposes of the university, not only efficiency or compliance. 

Staff and students can explore institutional guidance, AI literacy resources, examples from practice, workshops, communities of practice, research initiatives and collaborative projects related to AI in teaching, learning and assessment. 

This should probably not be a normal FAQ only. It should become a call-to-action block with links to: 

• guidelines 
• AI literacy resources 
• examples from practice 
• workshops/courses 
• research/collaboration opportunities