Incorporating AI
For SMU Instructors
Common assessment templates:
Incorporating AI into assessment requires balancing innovation with the integrity of learning outcomes. Start by identifying the core competencies your course aims to develop. Restrict AI use where it may replace the skills students are meant to learn.
When AI use is limited, be transparent about why. Clear rationale helps students understand how restrictions support their skill development and academic growth.
Issue Guidance on AI Use by Task
Decisions about generative AI should be made at the task level, not just the assessment level. This ensures AI use supports learning without replacing the core skills students are expected to develop.
The following section outlines sample constituent tasks for some common assessment types.
A written assignment completed outside class, requiring students to research, analyse, and present arguments on a given topic.
Typical Tasks Involved
- Ideation Draft outline
- Information/literature search
- Summarise documents
- Create first draft
- Edit for clarity and flow
- Feedback on work
Practical programming task where students design, implement, and test code to solve problems or build applications.
Typical Tasks Involved
- Understanding requirements
- Planning and design
- Code implementation
- Testing and debugging
- Documentation
- Deployment
A collection of structured exercises (often in math, science, or quantitative fields) requiring students to apply concepts to solve defined problems.
Typical Tasks Involved
- Problem comprehension
- Identify relevant concepts and formulas
- Solve problem
- Graphical / visual / tabular representation
- Check and verify solution
- Presentation and formatting
- Interpretation of results
Some factors to consider in determining whether to permit AI use:
- Core Skills: What core skills and knowledge do the assignment aim to develop?
- AI as a Tool: Can AI be used in ways that support, rather than replace, the intended skill development?
- Knowledge vs. Process: Is the assessment testing the student’s ability to derive the right answers or the process of reaching the right answers
- Complexity and Creativity: Will AI’s involvement risk oversimplifying the task, or can it enhance creativity and exploration?
- Current Guidelines: What are your current guidelines for AI use in assessment? Are they well-defined and clearly articulated to students?
- Tool Availability: Do all students have equitable access to AI tools? Would privilege access to certain tools give some students an advantage in the assessment?
- AI Skill Support: Have students been taught how to use AI tools responsibly and critically?
For each task within an assessment, instructors should specify whether to permit, permit with disclosure, or prohibit AI use.
- Permit: AI use is allowed, and instructors should state recommended ways for students to use AI for this task
- Permit with disclosure: AI may be used, but instructors should specify what students must disclose
- Prohibit: AI use is restricted, and instructors must design disincentives or prohibition measures.
The table below provides a sample of task-level guidance on AI use. Instructors may use a similar table to plan and issue guidance on use of AI with the provision of clear rationales.
| Task | Permitted | Permitted with disclosure | Prohibited | Details Instructor to provide further direction based on the level of AI usage indicated |
|---|---|---|---|---|
| Ideation | Set aside time for students to generate and discuss ideas in class without the use of AI; submit a summary at the end of the discussion. | |||
| Draft Outline | Students are required to start with a generated outline, which they then critique and improve. Disclosure through submitting the generated outline. | |||
| Information / literature search | Use Gen AI to search for relevant articles/publications (using tools such as ChatGPT's web search). |