Teaching & Learning Supports
The Use of Generative AI Tools
The CHTL has recently issued the advisory guidelines document in response to the University’s Principles for the use of generative AI tools in Teaching and Learning, and Assessment approved by the University Senate. To help students and colleagues understand what is expected of them regarding using generative AI tools in the courses, this webpage summarizes the key points and provides supplementary information for faculty members regarding how to integrate generative AI tools in courses and assignments to achieve the intended learning outcomes.
What are Generative AI Tools (e.g. ChatGPT)?
Artificial intelligence (AI) is “the capability of a machine to imitate intelligent human behavior” (Merriam-Webster). Generative AI technologies or tools (e.g. ChatGPT, Dall-E) leverage deep learning models to generate human-like “original” content, including audio, code, images, text, simulations, 3D objects, and videos. By utilizing natural language processing, many of these generative AI tools function as chatbots, simulating human-like conversations. This means that when a user submits a prompt, the chatbot generates content in real time as a response to the given prompt.
ChatGPT, developed by OpenAI, is one of the most popular generative AI tools. It is trained on large datasets, codes, and texts and draws upon this vast pool of data to generate responses. With its predictive technology, ChatGPT can produce or modify various textual works such as essays, computer code and reports.
While ChatGPT is well-known, there are many other generative AI applications that can also create images (e.g. Dall·E 2, Midjourney), videos (e.g. Synthesia, Pictory), music (e.g. Soundful, Mubert), code (e.g. Tabnine, GitHub), etc.
Generative AI has the potential to transform the operations of businesses (See AACSB) For example, text generation tools are commonly used in content creation, automated customer support, personalized messaging. Image and video generation tools are usually applied in product design, advertising, data augmentation, and customized training experience. With its ability to automate various tasks and generating insights for growth and innovation, generative AI will have substantial impact on the business landscape. You may refer to the McKinsey special report (June, 2023) for the economic and business potentials of generative AI in details. Therefore, it is essential for business school to integrate generative AI into classroom instruction and pedagogy to help our students.
Generative AI tools are also becoming increasingly relevant in higher education as they can help both instructors and students design and organize course materials, personalize course content, and enhance learning experience by generating original content in response to user prompts. It benefits learners in several ways, including:
- assisting with the writing process;
- summarizing and clarifying complex texts;
- providing grammatical and language assistance;
- encouraging discussion around rhetoric, style, and AI literacy;
- allowing instructors to focus on personalized assessments, on-going support, higher-quality feedback, with more efficient use of resources.
When integrating generative AI tools into business or courses, it is important to consider their limitations. Many of these limitations arise due to the way AI operates and the information available to them at the time of their training. They include:
- inability to comprehend the meaning behind their words or exhibit human-like levels of critical thinking;
- generating incomplete, inaccurate, or false information, include plagiarized text without proper attribution;
- generating biased output, as AI models may continue or intensify biases present in the training data.
The use of generative AI also raises several ethical concerns. The most widely discussed concern centered around education is the academic integrity issues. Others include:
- data privacy concern, as AI may collect or use personal data in training;
- potential misuse or abuse, as generative AI can be employed to produce false information, deepfakes, or other malicious content;
- equity and equality of access.
The University’s General Principles
Given the proliferation of generative AI tools and their utility in teaching and learning, as well as the society, the University has set out the following General Guiding Principles for the use of General AI in teaching and learning, and assessment.
The University’s General Guiding Principles
|Build and sustain human uniqueness|
Use of Generative AI in Assessment
|In alignment with the University’s guiding principles, the default approach is that students are ALLOWED and empowered to use generative AI tools in assessed assignment tasks.|
In some courses or some specific assessment tasks, instructors may restrict or prohibit the use of all or some generative AI tools for all or some assessment tasks. For example,
- accredited courses where the accreditation body has different guidelines or policies from the University;
- in-class examinations or tests;
- other assessment tasks in which the use of generative AI would prevent students from effectively demonstrating their achievement of the PILOs and CILOs.
It is crucial to communicate with students about the expectations for students’ proper and ethical uses of generative AI in the courses. Particularly, instructors are required to be transparent about the reasons relating to learning outcomes for using/prohibiting these tools for certain assessment tasks (if any).
Besides, they are encouraged to:
- clearly communicate appropriate and inappropriate use of generative AI tools to students.
- have open conversations about the potential benefits and drawbacks.
- discuss opportunities for generative AI to positively contribute to their discipline and prepare students for the future where AI will be embedded in many aspects of our lives.
- emphasize the importance of critical thinking and digital literacy.
- connect students to library resources for research and writing.
- encourage students to reflect on their learning process and engage with the course material actively to deepen their conceptual understanding.
Proper Use of Generative AI
Students are expected to use the generative AI tools available in the University library’s database, to ensure equality assess. Uses may include:
- explaining or clarifying concepts;
- demonstrating and guiding practices of techniques;
- planning and brainstorming on projects;
- giving feedback on drafts;
- generating samples for discussion and critical review.
However, students should NOT take the AI-generated content and present them as if they were their own work. Or else, it would be regarded as plagiarism (intentional or unintentional).
Unlike other traditional academic sources, there is no specific author for AI-generated content. Also, generative AI can be used in many other ways, such as refining writing, developing lines of argument or generating ideas. Therefore, it might be difficult to provide citations in traditional ways.
To uphold academic integrity, transparency and the ethical use of generative AI tools, any use of generative AI tools must be acknowledged by:
Students are required to complete and submit the following standard declaration.
I did knowingly use generative AI tools in this assignment. I followed the University’s guidelines for students on academic integrity. No content generated by generative AI tools has been presented as my own work. I take responsibility for the work submitted.
OR I did not use generative AI tools in this assignment.
Acknowledgement (if generative AI is used)
1. I acknowledge the use of [insert AI system(s) and link] to [specific use of generative AI]. The prompts used include [list of prompts]. The output from these prompts was used to [explain how the output was used].
The output from the above prompts was stored properly and will be submitted upon request.
- An example (including a few different uses) is provided for your reference.
- In-text Citations and Reference Lists:
In general, content produced by generative AI tools should not be considered reliable, accurate, or trustworthy representations of information. However, there may be times when the contents created by generative AI tools are included in the project submitted. In these cases, students are required to provide in-text citations for those material generated by generative AI and include them in the reference lists.
- Refer to https://apastyle.apa.org/blog/how-to-cite-chatgpt for citation and references guidelines for the APA style.
Checking for Improper Use of Generative AI Tools
If instructors suspect students of not following the University’s guidelines for students on academic integrity by knowingly presenting the output of generative AI tools as their own work, then the procedures governing student academic integrity apply.
Instructors are recommended to apply their professional judgement and use the following methods to check for improper use of generative AI tools:
- request students’ complete generative AI tools record of use.
- request earlier drafts of students’ work.
- request students to be orally examined on their submission.
IMPORTANT NOTE on the use of AI detection tools: AI detection tools, e.g. Turnitin AI detection, GPTZero, could be used to flag potential improper uses of generative AI tools. However, there is a lack of evidence regarding their effectiveness:
- There could be cases of false positive, i.e. incorrectly identifying fully human-written text as AI-generated text. For example, Turnitin admitted that there were higher-than-expected false positives1.
- It is unlikely for them to keep up, given the rapid pace at which AI is developed.
AI detection tools are NOT to be used as SOLE evidence of improper use of generative AI tools. Other methods, such as oral examination, must be used to verify cases of improper use.
Redesigning Assessment Tasks
In their recently published advisory guidelines which is accessible on BUniport (U-Wide Policies & Info -> Policies & Guidelines -> Quality Assurance, Teaching & Learning), the CHTL has proposed several general strategies to redesign assessments tasks so as to ensure assessments remain effective and focused on students’ demonstration of CILOs.
In order to help colleagues to redesign their assessment effectively, some practical suggestions are provided:
- diversify assessment methods and encourage creative expression using complex reasoning, graphics, mental maps, images, or videos to challenge the capabilities of AI.
- alter assessment process, such as staged submission of proposal, progress report, draft, and final version.
- ask specific questions that require logic or original analysis to make AI redundant.
- apply concepts to little-known cases and use less common or more recent topics to challenge AI.
- emphasize critical analysis over memorization of facts and figures.
- encourage students to think about the process and look at the subject from different perspectives.
- develop monitoring, curation, and critical thinking strategies to navigate the age of continuous and automated content.
- encourage students to review and evaluate AI answers to further their understanding of the subject.
- adopt authentic assessment with real-world or industry contexts, such as real-world case studies or competition, company-based or consultancy projects, work-integrated learning.
Integrating AI in Teaching & Learning
To integrate generative AI technology effectively, instructors should align it with course goals and learning outcomes. Teaching and learning activities, as well as assignments, that scaffold the process of learning may be well-suited to the integration of generative AI applications. When using generative AI, educators should create learning experiences that enable students to practice higher-order thinking skills, such as analyzing, evaluating, and creating. Some general suggestions include:
- Students can use ChatGPT and fact-check responses by finding primary and secondary sources to back up information.
- Students can generate a first draft using ChatGPT and track changes in a document to refine/edit.
- Reflect on prompt engineering and showcase how small changes can lead to major differences in output.
- Challenges students to explain the logic or reasons behind AI-generated responses.
- Students can compare the AI-generated responses with their own answers and evaluate if it might help improving their own answers.
- Require students to make a connection to or evaluate AI responses with the theories or knowledge acquired in class discussions.
- Simulate real-life tasks to facilitate contextual knowledge.
Also, as AI are used across business functions, teaching and learning activities might be designed to enhance students’ ability in applying generative AI in real business context. Some examples include:
|Business Functions||Examples of Applications|
|Marketing and Retailing||Crafting personalized marketing, social media and content creation (text, images and videos); analyzing customer feedback and sales data; improving sales force|
|HR and Operations||Assisting in screening and candidate assessment, providing self-serve HR functions, optimizing communication of employees, personalized training|
|Finance and Economics||Sentiment analysis, anomaly detection, personalized investment advice and financial services, data analytics, predictive modeling, algo-trading|
|Accountancy, Legal and Compliance||Enhancing auditability by tracking complex transactions, pulling data from vast amounts of legal documentation, drafting and reviewing annual reports, payables/receivables processing|
|Entrepreneurship, Innovation and R&D||Identifying new prospects and generating prototypes, personalization engines, predicting customer service issues, personalized marketing|
Generative AI technologies can also be used innovatively in business curriculum. It can be deployed to simulate real-life tasks to facilitate contextual knowledge. For example, ChatGPT can be:
- integrated into simulated negotiations, where students haggle with the chatbot playing the role of the opposing party for international trade or policy negotiation, bargaining or negotiation with labors, and so on.
- used as a sparring partner, with AI critiquing students’ ideas and asking for more explanation in their pitching of business proposal or marketing plan.
- used as a simulated stakeholder, with AI raising and enforcing students’ consideration in particular ESG issues.