Artificial Intelligence in Teaching & Learning

How We Think About AI in Teaching

Generative AI is becoming part of everyday work and learning. Our students will build careers in a world where AI tools are increasingly pervasive, and it is incumbent on us to teach them to handle these tools in an effective and discerning way. 

Instead of avoiding fast evolving emerging technologies like AI, we aim to understand and harness their potential. This includes adapting our teaching and assessment practices to prevent misuse.

We are committed to exploring how generative AI can be integrated into education. At the same time, we work to safeguard academic integrity and to maintain standards of academic rigour. Ensuring the responsible and ethical use of AI tools must be part of our response.

Our approach is to embrace the opportunities AI offers while ensuring that teaching and assessment continue to prioritise genuine student learning and sound academic judgement. The sections below outline our framework for AI use, assessment approaches, and examples from teaching practice.

We encourage instructors to experiment with AI in ways that improve efficiency, productivity, and teaching effectiveness - consider how AI can help address long-standing challenges, from scaling timely personalised feedback to enabling new forms of inquiry and collaboration.

For SMU Instructors: If you are currently using or planning to use AI for formative feedback or other pedagogical experimentation, please email CTE at cte@smu.edu.sg. CTE will then contact you to provide support and work with you. The objective is to learn from your experience and better understand the benefits and challenges of such use.

AI systems vary in accuracy and reliability, and their use raises important questions about academic integrity and student learning. Human academic judgement must remain central, especially in high-stakes assessment. 

A working group comprising faculty and representatives from CTE and IITS is testing the effectiveness of different AI grading tools. The recommendations of the group will inform further policy changes. 

Beyond issues of integrity, over-reliance on AI can limit the development of knowledge, critical thinking, and independence. 

This challenges instructors to redesign their courses and assessments in a way that meaningfully engages students in the learning process rather than allowing key aspects of thinking to be outsourced.

Framework for the Use of GenAI Tools
 

Use of Generative AI Tools by Students as a Learning Aid

Generative AI tools have the potential to enhance students’ understanding and knowledge leading to better learning outcomes. Nevertheless, students need to be made aware of the limitations and ethical use of such tools.

The Student Success Centre has launched the Guide to Learning with AI online module to educate students on how generative AI tools can be used as a learning aid in an effective and responsible manner. This module can be found on our learning platform.

 

Use of Generative AI Tools by Instructors as a Pedagogical Tool

Generative AI tools can lead to improved teaching quality and support instructors to create educational content. A series of webinars / workshops has been planned to provide support to instructors on the effective use of generative AI tools in teaching, including demonstrating best practices for incorporating these tools in their teaching.

Generative AI tools can also help improve your productivity as an instructor. For example, you may: 

  • Use it to assist in the creation of educational content, such as practice exercises, interactive lessons, or rubrics for your assessments.

  • Use it to seek ideas for assessment questions formulation.

     

Generative AI tools have raised new questions about how student learning should be assessed. Instructors are understandably concerned that students may use these tools to produce work that is presented as their own. At the same time, these technologies are becoming part of professional practice across many fields.

Our goal is therefore not only to prevent misuse, but also to ensure that assessments continue to support meaningful learning and the development of critical thinking.

Read about the 'AID' framework under 'Our Approach to Assessment' below.

🔒 For SMU Instructors

Instructors seeking further guidance on the use of AI tools for teaching and learning are welcome to contact the Centre for Teaching Excellence.

The CTE team provides recommendations and advice tailored to specific course contexts and pedagogical goals.

Reach out to cte@smu.edu.sg for more information.

Our Approach to Assessment - AID

SMU approaches AI in assessment through three complementary strategies: Adapt, Incorporate, and Detect.

Adapt assessment design

Instructors may choose to redesign assessments so that they emphasise forms of work that are less easily generated by AI tools. This could include:

  • designing questions that require personal reflection, applied analysis, or discipline-specific reasoning

  • incorporating staged or process-based assignments

  • using formats such as presentations, discussions, or in-class work alongside written submissions

Adapting assessment helps ensure that student work reflects genuine understanding and engagement.

 

Incorporate AI where it supports learning

In some cases, allowing students to use generative AI tools can enhance learning. When AI use is permitted, instructors should:

  • clearly state what kinds of use are allowed

  • provide guidance on how the tools should be used

  • require students to acknowledge and credit AI use where appropriate (e.g., following APA or Chicago citation guidance)

Integrating AI thoughtfully can help students learn to use these tools critically and responsibly.

 

INCORPORATE AI INTO YOUR ASSESSMENTS >

Detect potential misuse

Where misuse is suspected, instructors may make use of available detection tools and follow established university procedures.

SMU instructors should follow the DRIVE approach when responding to suspected unauthorised use of generative AI in assessments.

 

DETECT MISUSE WITH THE DRIVE APPROACH >

🔒 For SMU Instructors

For assessments where the use of generative AI tools is permitted, students should be told the specific areas and tasks where generative AI tools could be applied and what constitutes unauthorised use. In this link (Statement Templates), instructors can access templates of statements they can adapt according to their requirements:

  1. Statement to explain the use of Generative AI tools in assessment to students.

  2. Statements of Declaration for students in assessments where the use of Generative AI is permitted, restricted, partially permitted under certain conditions specified by the instructor.

    In the event that unauthorised use of Generative AI tools is suspected or detected, please apply the DRIVE approach and refer to the Faculty Handbook for Academic Violation.

    Instructors may seek guidance from Associate Deans, and also contact the University Council of Student Conduct for more information:

    Tel : 6828 0885
    Email : studentconduct@smu.edu.sg
    Website : https://smu.sg/UCSC

Communicating Expectations to Students

Clear communication is essential. Instructors should explain their policies on AI use at the start of a course and ensure students understand both the expectations and the rationale behind them.

This includes:

  • clarifying when AI use is permitted and when it is not
  • explaining what counts as appropriate acknowledgement of AI use
  • reminding students that unauthorised use may constitute a violation of the SMU Code of Academic Integrity

Teaching Students to Use AI Well

Beyond setting boundaries, instructors are encouraged to help students develop thoughtful and responsible ways of working with AI tools. This may include:

  • discussing the strengths and limitations of AI-generated responses
  • encouraging students to verify and critique AI outputs through further research
  • asking students to reflect on how they used AI in completing a task
  • allowing students to explore questions using AI as a starting point for deeper inquiry

These practices help students develop the judgement and critical thinking they will need in a world where AI tools are increasingly common.

 

Use Cases of AI in Teaching

 

Many leading AI tools such as ChatGPT and Gemini provide users with out-of-the-box means of customising chatbots for specific tasks and uses. This is accomplished by writing a set of instructions, which guide the interaction with the user, and establishing a knowledge base with documents and information that the chatbot can refer to in generating its responses.

Custom chatbots are highly flexible tools, with uses that depend on the instructions given. Some ideas for use in your teaching include:

  • Offering students personalised and immediate feedback on a formative task

  • A thinking partner that helps students develop better arguments

  • Interactive exercises such as teaching the chatbot as if it were a novice learner, helping students to reinforce their own learning

Today’s workplace demands professionals who can work effectively alongside AI, integrate it into business processes, and apply it responsibly and ethically. As AI becomes embedded in everyday professional practice, proficiency in using these tools is no longer optional - it is a core competency. Educators have a responsibility to redesign courses and assessments to prepare students for a workplace where the ability to use AI thoughtfully and strategically is essential.

Integrating AI is not about chasing a trend. It is about aligning teaching with reality. The workplace has changed, and responsible educators must change with it. This journey requires constant recalibration. It involves redesigning assessments, rethinking rubrics, and confronting issues on over-reliance and fairness. Here are some practical pointers to help you with this process of assessment redesign:

  • Start with learning outcomes: Clarify what students must know and be able to do. These outcomes should remain valid even in an AI-augmented workplace (e.g., critical thinking, ethical judgment, disciplinary reasoning, problem-solving). Determine how AI can support (not replace) those outcomes.
  • Redesign rubrics to assess higher-order competencies: As AI tools increasingly assist with drafting, formatting, coding, or analysis, assessment must emphasise what students uniquely contribute: interpretation, synthesis, contextual judgment, strategic decision-making, creativity, and ethical reasoning. Rubrics across the course should reflect this shift.
  • Establish clear AI use policies: Provide transparent guidance on acceptable and unacceptable AI use. Require students to declare where and how AI tools were used, including prompts where appropriate.
  • Diversify assessment modalities: Avoid relying solely on take-home, AI-compatible outputs. Incorporate varied forms of assessment such as oral components, live demonstrations, applied simulations, reflective commentaries, and process documentation. Multiple modalities provide a more complete picture of student learning and reduce over-reliance on AI-generated responses.

  • Embed AI stress-testing into course design: Before implementation, test assessment tasks using current AI tools to evaluate whether they can be completed without meaningful human input. AI integration is not a one-time redesign. Revisit this process periodically, as AI capabilities evolve rapidly. This ensures tasks remain rigorous and purposeful.


As students prepare to enter a world where generative AI is embedded in everyday professional practice, the distinct skills they contribute shift from merely producing responses to exercising critical evaluation and sound judgment.

Integrating critical thinking into the classroom is essential to ensure that students develop the judgment, discernment, and intellectual independence needed to navigate an information-rich and AI-assisted world. If you are designing or redesigning your course to integrate critical judgment in the age of AI, consider the following strategies:

  • Use AI as a thinking stimulus, not a solution: Have students critique, extend, or challenge AI-generated ideas in small-group discussions to surface reasoning and assumptions.
  • Make verification a classroom routine: Incorporate activities where students compare AI responses with scholarly sources, data sets, or alternative perspectives to identify gaps and inconsistencies.
  • Surface the thinking process: Build in think-aloud exercises, peer explanation, or collaborative annotation of AI outputs so students articulate how they evaluate claims.
  • Create moments of productive friction: Present conflicting AI outputs or ambiguous cases that require students to debate credibility, bias, and contextual appropriateness.

Designing with these principles in mind helps position AI as a tool for sharpening judgment rather than substituting for thought.

Well-designed questions play an important role in helping students learn. They allow students to test their understanding of key concepts and to identify gaps in their knowledge. Incorporating quizzes can make classes more engaging for learners.

However, creating high-quality questions can be challenging. While many instructors would like to provide students with more opportunities for practice, they are often constrained by limitations in time and energy.

Generative AI can help instructors: 

  • Create questions more efficiently
  • Design quizzes that include questions targeting different levels of thinking based on Bloom's Taxonomy
  • By providing inspiration and increasing the diversity of questions

Apart from tools specifically designed for question generation such as UseMemo, similar outcomes can be achieved with other generative AI platforms. You may find the following resources useful for designing prompts for this purpose.

Video content has become an important part of many students' learning experiences. However, high-quality videos can be tedious and challenging for instructors to create, requiring the right equipment, a suitable recording environment, and a considerable time investment. In many cases, instructors must record multiple takes to achieve a polished result and need to re-record the videos to update their content.

AI avatar tools such as HeyGen can be powerful enablers for instructors who may not have prior video production experience or access to specialised recording equipment. Compared with traditional recording workflows, they offer clear advantages in terms of efficiency and production quality, while also making it much easier for instructors to update or revise their video content when course materials change. When used thoughtfully, these tools can still preserve the instructor's voice, helping students feel connected even when learning asynchronously.

Here are some ideas for getting started with HeyGen (or other AI avatar video creation tools): 

  • Start small, such as by creating a video introducing your research, to build familiarity and confidence with HeyGen's capabilities.
  • Focus on creating videos for key concepts or difficult topics in your course and avoid using them excessively.
  • In general, keep these videos brief and to the point to better maintain student engagement.