Generative AI at the U of Regina

Artificial intelligence (AI) refers to the simulation of human intelligence in machines, enabling them to perform tasks that typically require human-like thinking. Within this realm, Generative AI stands out as a subset that can create new content and artefacts, such as text, images, audio, and video, based on patterns learned from large amounts of data. Unlike other types of AI systems that merely analyze and interpret data, generative AI produces novel and unique outputs from simple text prompts. The recent emergence and popularization of powerful generative AI platforms like ChatGPT offers both promising opportunities and new challenges for university teaching and learning, particularly in areas related to academic integrity, intellectual property, and creativity.

ChatGPT is a powerful, conversational generative AI system (i.e., chatbot) developed by OpenAI that was launched in November 2022. ChatGPT utilizes deep learning techniques to comprehend and generate human-like responses to the input it receives. ChatGPT’s release to the general public has led to concern in both K-12 and post-secondary learning institutions about the tool’s potential use for student plagiarism and other academic misconduct. In considering the uses of generative AI in teaching and learning, it is important to acknowledge that ChatGPT is only one of many similar tools that are available for public use (e.g., Google’s Bard, Anthropic’s Claude, Microsoft’s Bing, Meta’s LLaMA, Stanford’s Alpaca). Additionally, based on announcements throughout the tech industry, generative AI systems will soon be ubiquitous and will be embedded within commonly used tools such as Microsoft 365 and Google Workspace.

In post-secondary education, generative AI has potential uses across teaching, learning, research, and administration. Some examples specific to teaching and learning might include:

• Assisting in the development of course learning outcomes, syllabi, assignments, and rubrics.
• Producing personalized learning content such as tailored explanations, outlines, dialogues, lectures, and study aids.
• Generating accessible course materials through automated transcriptions, image analysis, language translation, and accessible document generation.
• Providing adaptive learning assistance via AI tutoring.
• Supporting formative assessment techniques through automated, on-demand feedback.

As generative AI becomes more ubiquitous and accessible, it is expected that its applications in teaching and learning will become more commonplace and varied.

The decision of whether or not to allow the use of ChatGPT or other generative AI in academic work ultimately rests with each instructor, based on their pedagogical and disciplinary expertise. However, instructors have a crucial role to play in mitigating potential misuse. It is recommended that instructors take the following actions:

• Guidelines – It is important for instructors to clearly lay out the rules, boundaries, and expectations for using generative AI in coursework and other academic activities. It is strongly recommended that instructors incorporate a statement on the use of generative AI into course syllabi. Please consult the University of Regina’s sample syllabus statements regarding the use of generative AI in coursework.
• Understanding – It is essential for instructors and students to better understand the types, capabilities, and potential uses and misuses of generative AI. Instructors should familiarize themselves with these technologies and work to foster student conversations and explorations regarding these emerging tools.
• Identification – Instructors should identify innovative ways to incorporate generative AI into curriculum, such as brainstorming ideas, producing code examples, generating historical dialogues, assisting with research, etc. Concurrently, instructors should work to identify the constraints of using tools like ChatGPT within the scope of their subject matter, emphasizing limitations such as the risk of generating outdated data or incorrect information (including fabricated citations), the possibility of producing analyses based on inherent biases, or the risk of using language or references that may be culturally insensitive or inappropriate.
• Disclosure – Instructors should require students to disclose when and how they’ve incorporated AI into their coursework. This promotes transparency and academic honesty. Instructors may want to provide a sample statement that students can adapt in order to clearly communicate how they have used AI in their work; an example of such a statement (which was generated with the help of ChatGPT) might be: “This [text, image, or video] was produced by the author using assistance from [insert generative AI provider]. The author affirms that they have thoroughly examined, modified, and refined the initial AI-generated draft and acknowledges that they are accountable for the content of this assignment.” Additionally, the APA, MLA, and Chicago citation guides have released information on how to cite AI-generated content in academic work. APA and Chicago currently recommend citing such content as “personal communication,” while the MLA style guide includes more detailed instructions. It is equally important for instructors to exemplify these ethical principles by transparently disclosing when they’ve used AI or other related tools in coursework.
• Ethics – Instructors should foster open dialogue that highlights the ethical considerations when using generative AI. Such conversations might include issues of bias, representation, copyright, and authorship. Instructors should also endeavour to protect the privacy and intellectual property of students in any use of generative AI in teaching, including in assessment and feedback.

Tools designed to detect AI-generated content have demonstrated varying degrees of reliability; therefore, such tools should not be relied upon to ensure academic integrity. At present, Turnitin, the University of Regina’s institutionally adopted plagiarism-detection tool, produces an AI-detection report alongside the originality report when used for text-based assignments. This additional report is only available to the instructor. Due to the unreliability of AI-detection tools, and the possibility of false positives, the AI-detection report cannot be considered conclusive proof of academic misconduct; however, the report may be presented to the investigating dean alongside other potential evidence of cheating (see below for other possible signs). It should also be noted that Turnitin is presently the only institutionally adopted plagiarism- and AI-detection tool and that instructors are not permitted to use non-approved tools for the purpose of AI-detection.

In cases where instructors suspect that student coursework has been partially or wholly AI-generated without proper disclosure/citation, instructors are encouraged to look for additional signs, such as:

• Repetitiveness (overuse of certain words, phrases, or sentences)
• Lack of nuance or depth (generalized or overly broad statements)
• Lack of personal experience (generic descriptions of feelings or experiences)
• Inconsistencies (contradictory statements within the text)
• Neutral tone (lacking a distinct voice or style)
• Semantic errors (subtle mistakes in language or word choice)
• Lack of cultural or contextual awareness (limited knowledge of specific cultural contexts)
• Absence of critical thinking and analysis (missing nuanced or critical perspectives)
• Incorrect citations (including entirely fabricated citations)
• Generic content (lack of references to course lectures and materials)
• Poor transitions (sections of text on different topics with little continuity)

Note that many of the issues listed above may also point to other types of academic misconduct (including plagiarism and contract cheating) or to poor writing ability.

If you’d like to receive support for the use or integration of generative AI in teaching and learning, feel free to contact the Centre for Teaching and Learning via email at ctl@uregina.ca.

If you require support regarding a matter related to academic misconduct, please consult with your investigating dean and/or consult the instructor’s page on the Academic Integrity Hub.