Generative AI, also called Gen-AI is a type of artificial intelligence that allows machines to generate new content based on a series of prompts entered by its user. Outputs can include text, images, videos, sound, code and other media.
This guide provides resources to help you understand Gen-AI, its limitations, and how to use it ethically in your studies.
Training
Gen-AI systems are trained on large data sets, learning patterns and features from existing data.
Input and Output
Gen-AI systems are often prompted with natural language from a human, which can be text or voice, or copied text from elsewhere. From these inputs (prompts) Gen-AI systems generate new content that resembles what they have learned. This could be text, images, videos, music, or code.
Examples of Gen-AI
Chatbots (ie Large Language Models or LLMs): ChatGPT, Copilot, Gemini, and LLaMA
Text-to-image systems: Stable Diffusion, Midjourney, and DALL-E
Text-to-video generators: Sora
Not all Gen-AI is equal
Different Gen-AI models have varying capabilities depending on their training data, architecture, and specific design, leading to differences in output quality, accuracy, and suitability for different tasks, meaning some Gen AI models will perform better than others depending on the situation.
Why Gen-AI varies
Training data diversity: The quality and variety of data used to train a Gen-AI model significantly impacts its output.
Model complexity: Different architectures and algorithms used in building Gen-AI models can result in varying levels of sophistication and performance.
Bias in data: If the training data contains biases, the generated outputs will reflect those biases. Like any source the output of Gen-AI needs to be scrutinised and evaluated.
Application specific: Some Gen-AI models are optimized for specific tasks like text generation, image creation, or code writing, making them less suitable for other tasks.
Basic terminology
Further reading
Policies and Procedures
Student Support Guides
Citing Gen-AI
When approved to use Gen-AI in assessments, adhere to the relevant referencing guidelines for advice on how to cite the use of Gen-AI in your in-text and end-text references.
Click on your school's referencing style guide to see the rules relating to citing approved use of Gen-AI use in your assignments.
Referencing Guides:
AGLC | APA | Chicago | AMA (Vancouver)
Statement of acknowledgement
A written statement of acknowledgement is also required when using Gen-AI in your assignments. Your course coordinator will provide guidance on how to acknowledge your use of Gen-AI in your assessment.
In your acknowledgement you should provide:
A written statement acknowledging the use of Gen-AI
Specify what Gen-AI tools and technology were used
Include a list of prompts used
Explain how the outputs were used in your work
For example:
I acknowledge the use of [insert name of AI tool] to [insert description of usage]. The prompts used were [insert list of prompts]. The outputs generated from these prompts were used to XXX.
Gen-AI tools
If using any Gen-AI tools you should:
Gen-AI can:
Types of biases in content generated by Gen-AI include:
To try and mitigate AI hallucinations you should use an AI tool designed for the purpose you intend to use it for and provide clear, specific direction in your prompts.
Refer to Crafting useful prompts for further advice.
Most of the free Gen-AI tools don't provide the sources used to generate the response, making it difficult to fact check the validity of the outputs.
If a reference is provided this could be made up or a mash up from different sources to create a fake reference.
You will need to do your own research to ensure the outputs you receive are factually correct
There is no reassurance that your uploaded data is protected. There is potential for your data to be collected and on-sold.
The carbon footprint of Gen-AI models is significant. Data centres used to develop and deploy Gen-AI technologies consume massive amounts of energy and produce high levels of carbon emission.
Open Educational Resources (OERs)
While OERs can be valuable learning materials, they may not always align with Notre Dame's policies or academic standards. Always cross-check information with the University's current policies and guidelines for using Gen-AI in assignments or research. What is permitted at one institution may not be allowed at another.
Credo
Sage Campus
Linked-in Learning
LinkedIn Learning provides access to an on demand library of videos developed by industry professionals on everything from essential software skills in Excel, Word, PowerPoint, Teams, SharePoint and more through to educational technology, business analysis, project management, leadership, professional and career development
Copyright © 2025 The University of Notre Dame Australia | CRICOS Provider Code: 01032F | TEQSA PRV12170 | RTO Code 0064