SMU’s stand
SMU does not permit the use of AI to grade summative assessments.
This is because summative grading has direct implications for learners’ progression, awards and fairness. Current AI tools may not reliably recognise nuance, context, originality or discipline-specific reasoning. They may also introduce bias, favour conventional responses, or produce inconsistent evaluations.
Accordingly, final grading decisions for summative assessments must remain the responsibility of the instructor.
Limited use for formative feedback and pedagogical experimentation
AI may be used in limited contexts for formative feedback or pedagogical experimentation, where it provides clear value-add to learners and does not replace instructor judgement.
Examples may include using AI to support:
- rapid feedback on draft work;
- personalised feedback on specific areas for improvement;
- feedback on low-stakes practice exercises;
- generation of indicative comments for instructor review;
- experimentation with new approaches to formative learning support.
In such cases, AI should be used to augment, not substitute, the instructor’s role. Instructors remain responsible for ensuring that feedback is appropriate, accurate, fair and aligned with the intended learning outcomes.
Faculty who are using or experimenting with AI for formative feedback or grading-related purposes should inform CTE, so that emerging practices can be better understood and supported.
Work in progress
SMU’s working group on AI-assisted grading and feedback is finalising its recommendations on the appropriate use of such tools, including good practices and protocols for the university. These recommendations will inform future refinements to SMU’s guidance, to ensure that any use of AI-assisted feedback or grading is accurate, equitable, pedagogically meaningful, and aligned with SMU’s assessment principles.