Welcome to GenAI Literacy & Numeracy
This guide has been designed to provide foundational knowledge and practical skills for using Generative AI (GenAI) tools such as ChatGPT, Microsoft Copilot, and Google Gemini effectively, ethically, and responsibly throughout your studies and into your future career.
Learning to use these tools represents an essential digital literacy skill for the modern tertiary education environment. Whilst GenAI offers considerable opportunities to enhance learning outcomes, it also carries significant responsibilities. Understanding how to navigate this technology remains crucial for maintaining academic integrity and maximising the value of your education.
Upon completion of this module, you will be able to:
- Understand how Large Language Models function and their limitations
- Apply effective prompt engineering techniques in your discipline
- Use GenAI tools ethically whilst maintaining academic integrity
- Use the SAGE Anchor, Check, Defend model to verify AI responses and prepare to explain your work
- Recognise when and how to properly acknowledge AI assistance
Module Structure
Lesson 1: What is Generative AI?
Before any tool can be used effectively, it is essential to understand what it is, how it functions, and what its limitations are. This foundational knowledge forms the basis for responsible and effective GenAI usage.
Defining Generative AI
Generative AI Generative AI refers to artificial intelligence systems capable of creating new, original content including text, images, audio, video, or computer code based on patterns learned from training data. refers to a category of artificial intelligence capable of creating new and original content, including text, images, audio, video, or computer code. The most widely recognised examples include:
A conversational AI that generates human-like text responses
Another powerful conversational AI chatbot platform
A GenAI assistant integrated into Microsoft tools and Bing search
AI systems that generate images from text descriptions
These tools are powered by complex algorithms known as Large Language Models (LLMs) LLMs are sophisticated pattern-recognition systems trained on vast amounts of text data from the internet. They predict the most likely sequence of words based on learned patterns rather than "understanding" content in the human sense. .
How Do LLMs Actually Work?
LLMs do not "think" or "understand" in the manner that humans do. They are incredibly sophisticated pattern-recognition systems that have been trained on vast amounts of text and data from the internet.
When a prompt is provided to an LLM Large Language Model - an AI system trained to predict and generate text by learning patterns from massive datasets. , it does not search for a direct answer. Instead, it statistically predicts the most likely sequence of words to follow, based on patterns identified within its training data. This can be conceptualised as an exceptionally advanced version of predictive text functionality found on mobile devices.
💡 Simple Analogy
Imagine typing on your phone and it suggests the next word. An LLM works similarly, but on a massively more complex scale, predicting entire sentences, paragraphs, and documents based on patterns it has learned from billions of text examples.
Why do answers sometimes change? Randomness
Randomness settings (often called “temperature”) control how varied answers are.
- Higher temperature → more creative; answers may differ each time.
- Lower temperature → steadier; answers are more consistent.
Simple rule: for study plans or assessment prep, prefer lower temperature to keep outputs consistent with your instructions.
A simple instruction recipe Prompt structure
- Role: “You are a study tutor…”
- Task: “Explain X in ~120 words…”
- Limits: “Use plain English; short paragraphs.”
- Quality check: “Define two key terms and give one example.”
- Optional example: Paste a tiny sample to copy the style.
Clear, short steps help the tool stay on task.
Beyond text: working with multimedia Images / PDFs
Example: ask an AI tool to summarise a PDF page, describe a diagram, or explain a chart from your notes.
Always compare the explanation with your unit materials, lecture notes, and verified sources before relying on it.
Tip: Ask for a brief explanation of a diagram in your notes—then check it against your materials.
Watch out for fabrications (“hallucinations”) Stay critical
Fabrications happen when the tool makes things up (e.g., fake facts or fake references) that look real.
- Never copy a reference from the tool without finding it yourself.
- Check claims using library databases, textbooks, or your unit site.
If AI output looks confident, that does not mean it is correct.
Key Capabilities and Limitations
Understanding what GenAI can and cannot do represents the first step towards using it responsibly. The following table outlines these key considerations:
| ✓ Capabilities (What GenAI does well) | ⚠️ Limitations (Where caution is required) |
|---|---|
|
Brainstorming and Generating Ideas
Can serve as an effective starting point for essays, projects, or creative work by suggesting multiple angles or approaches. |
"Hallucinations"
AI hallucinations occur when the model generates information that appears plausible but is entirely fabricated, including fake statistics, citations, or facts.
and Factual Errors
GenAI can invent facts, statistics, and even academic references that appear legitimate but are completely fabricated. |
|
Summarising Complex Information
Can assist in breaking down dense readings or difficult concepts into more accessible formats. |
Bias
Due to training on internet data, GenAI can reproduce and amplify existing societal biases related to race, gender, and other factors. |
|
Improving Language and Grammar
Can function as a sophisticated grammar checker to enhance the clarity and accuracy of written work. |
Lack of Critical Thinking
AI cannot analyse, evaluate, or form original arguments. It merely reproduces learned patterns. Critical thinking remains a human responsibility. |
|
Creating Study Plans
Can assist in organising revision schedules, creating practice questions, or structuring learning activities. |
Data Privacy Risks
Public GenAI tools often store conversations to further train their models. Personal, sensitive, or confidential information should never be entered. |
GenAI is a powerful tool, but it is not a source of truth. You remain responsible at all times for the final work submitted under your name.
Lesson 2: How to Talk to AI
The quality of output received from a GenAI tool depends almost entirely upon the quality of instructions—or "prompts A prompt is the instruction or question provided to an AI system. Effective prompts are clear, specific, contextual, and provide sufficient detail to guide the AI toward useful outputs. "—provided to it. Learning to construct effective prompts represents the key to unlocking the tool's potential as a learning assistant.
From Simple Instructions to Effective Prompts
Prompting can be conceptualised as a conversation. A vague question yields a vague answer, whilst a detailed, specific prompt generates a substantially more useful response. The following techniques demonstrate how to improve prompting effectiveness.
1. Instruction Prompting (Be Clear and Specific)
This represents the most fundamental level of prompting. Rather than posing a simple question, it is necessary to provide context, constraints, and a desired format.
Example: Climate Change Essay
2. Role Prompting (Give the AI a Job)
This represents a powerful technique wherein the AI is instructed to adopt a specific persona. This approach assists in framing responses and provides more targeted outputs tailored to particular contexts or disciplines.
Role prompting can be adapted to any field of study. Some additional examples include:
- The Tutor: "Act as a university tutor for a first-year biology student..."
- The Debate Partner: "You are my debate partner. Challenge my position with strong counter-arguments..."
- The Skeptical Editor: "Act as a skeptical editor. Identify weak claims and logical fallacies..."
3. Few-Shot Prompting (Provide Examples)
When output in a specific style or format is required, providing one or two examples for the AI to follow can significantly improve results.
Example: Creating Summaries
Prompt:
"I will provide a complex academic paragraph, and your task is to summarise it into a single, clear sentence.
Example Paragraph: 'The proliferation of digital communication technologies has fundamentally altered the socio-political landscape, enabling unprecedented levels of civic mobilisation whilst simultaneously creating new vectors for misinformation and polarisation, thereby presenting a complex duality for modern democratic systems.'
Example Summary: 'Digital technologies have made it easier for people to organise for political causes, but they have also increased the spread of fake news and division.'
Now, please provide a one-sentence summary for the following paragraph: [Insert your paragraph here]."
Effective prompting is an iterative process. Begin with a clear prompt, evaluate the response, and then refine your instructions to improve subsequent results. It represents a dialogue rather than a single command.
Lesson 3: Using AI the Right Way
Using GenAI at university extends beyond merely obtaining satisfactory results; it concerns using the technology with integrity. Your degree represents a testament to your skills and knowledge, not those of an AI system. This lesson addresses the most important principles that must be followed.
This represents the most critical rule of all. At CQUniversity, permission to use GenAI is not universal. It is determined at the unit and assessment level.
- Always consult your unit profile and specific assessment instructions first. Your Unit coordinator will specify whether AI is permitted, prohibited, or subject to limitations for a particular task.
- If uncertainty exists, always seek clarification from your tutor or unit coordinator. Do not make assumptions.
Submitting AI-generated work when it is not permitted constitutes a serious breach of academic integrity and will be treated accordingly.
The Do's and Don'ts of GenAI Use
The following guide outlines responsible GenAI usage, based upon best practices from universities across Australia:
DO utilise AI for tasks such as brainstorming, initial research, and understanding difficult concepts. These activities support your learning process.
DO verify whether AI is permitted for a specific assessment by consulting your course documentation before use.
DO fact-check all claims against reliable academic sources from the CQUniversity Library. Never accept AI output at face value.
DO follow specific citation guidelines provided by your course or the university when AI assistance is permitted.
DO accept complete responsibility for your submitted work. Any errors, biases, or fabrications from the AI remain your responsibility.
DO consider how AI tools can support your development of critical thinking, analysis, and writing skills rather than replacing them.
DO save links to your chats or export your prompt history. If questioned about your work, being able to show your specific "conversation" with the AI is your best defence.
DON'T copy and paste AI-generated text directly into your assignment and claim it as your own work. This constitutes plagiarism.
DON'T use AI to generate entire sentences, paragraphs, or complete assignments. The work submitted must be genuinely your own.
DON'T rely on AI for academic references. It frequently invents "hallucinated references" that appear legitimate but are fabricated. Locate and cite your own sources.
DON'T enter personal, sensitive, or confidential information into public AI tools. Use university-endorsed tools where available.
DON'T use AI in ways that prevent you from developing essential skills such as critical thinking, analysis, and academic writing.
DON'T assume that AI use is permitted unless explicitly stated in your assessment guidelines. When in doubt, ask.
DON'T delete your chat logs. Without them, you may struggle to prove that you used the tool responsibly if asked to demonstrate your process.
How to Acknowledge and Cite GenAI
If permission has been granted to use GenAI in an assessment, transparency about its use is required. Honesty represents a core value of academic integrity. Whilst your unit coordinator may provide specific instructions, a common approach involves:
- An in-text citation where the AI-generated idea is utilised
- A reference list entry detailing the tool employed
- A declaration or appendix describing which tool was used, the date of use, the prompts employed, and how it contributed to the final work
Example Citation (APA 7th Style):
In-text citation:
Reference list entry:
Always consult the CQUniversity Library guide for the most current citation requirements.
The do's and don'ts in this lesson are the everyday version of a formal policy. The SAGE Assessment Policy sets out the full rules for AI use in assessment — what counts as AI use, how to disclose it, and what happens if you cannot defend your work.
It also provides ready-to-use declaration templates you can adapt when you acknowledge AI assistance. Remember that your unit profile and assessment instructions always take precedence.
If Your Work Is Wrongly Flagged as AI-Generated
AI-detection tools are not reliable. They produce false positives, and research has shown they disproportionately misflag work by students who write in English as an additional language. A "detector" score is not proof, and most universities do not treat it as standalone evidence. Even so, you may occasionally be asked to account for your work — so it is worth knowing how to protect yourself.
The most effective way to show that work is your own is to be able to show how you produced it:
- Keep your drafts and version history — Word, Google Docs, and OneDrive all retain edit history that shows your work developing over time.
- Save your notes, outlines, and research — the materials behind your work are strong evidence of authorship.
- Keep your AI chat logs — if you used AI in a permitted way, the conversation shows exactly what you asked for and what you did with it.
- Be able to explain your work — if you can talk through your reasoning, sources, and choices, you can demonstrate ownership in person.
Stay calm — being asked is not the same as being accused. Provide your drafts, notes, and any AI chat logs, and offer to talk through your process. You have the right to explain your work. The SAGE Assessment Policy sets out how this is handled, including what happens if AI-assisted work cannot be fully defended.
Your Learning is the Priority
The ultimate objective of your university education is the development of knowledge and skills. Over-reliance on AI can impede that process. Consider these critical questions when using GenAI tools:
- 🤔 Am I using this tool to support my thinking or to replace it?
- 📚 What skills am I failing to practice by outsourcing this task to AI?
- 🎯 How can I use this tool to enhance my learning, not merely to obtain an answer?
This is an evolving area. University policies on AI-generated visuals are still being developed. General guidance:
- Check your unit profile first – Some assessments prohibit AI-generated visuals
- Graphs from your own data: Using AI to visualize data you collected/analyzed is generally acceptable (like using Excel)
- AI-generated images/diagrams: Policies vary widely. Always ask your unit coordinator
- Always acknowledge: "Figure created using [tool name] on [date]"
- Never use AI images that misrepresent reality (e.g., fake data visualizations, misleading photos)
When in doubt, ask. This technology is changing faster than policies can keep up.
Your degree represents more than a credential—it reflects the skills, knowledge, and critical thinking capabilities you have developed. GenAI should enhance this development, never replace it.
Lesson 4: Practical Applications
Whilst the preceding lessons have established foundational principles for using GenAI responsibly, it is equally important to understand the diverse ways these tools can be applied throughout various stages of the learning process. This section draws upon evidence-based practices from Australian universities to demonstrate how GenAI can serve as a study companion when used appropriately.
All applications described below are contingent upon permission being granted in your specific unit and assessment. Always consult your unit profile before employing any of these strategies, and ensure that GenAI use enhances rather than replaces your learning.
1. Study Planning and Organisation
A well-structured study plan represents a cornerstone of academic success. GenAI tools can assist in creating personalised study schedules, recommending time management strategies, and helping to balance competing priorities.
2. Understanding Complex Concepts
GenAI can function as a virtual tutor to assist in comprehending difficult concepts by providing explanations at appropriate levels, drawing connections between related ideas, and offering multiple perspectives on the same topic.
3. Note-Taking and Information Organisation
Effective note-taking represents a critical academic skill. GenAI can suggest various note-taking methodologies, create templates for organising information, and help structure complex material into comprehensible formats.
4. Assessment Preparation and Planning
GenAI can assist in the preliminary stages of assessment preparation through brainstorming, structural planning, and critical questioning—whilst remembering that the final work must remain authentically your own.
5. Improving Writing Clarity and Style
Once initial drafts have been created through your own effort, GenAI can provide feedback on clarity, coherence, and structure—functioning similarly to a writing centre consultation.
6. Visual Content and Data Representation
Visual representations can enhance how you communicate data and findings. GenAI tools can suggest appropriate chart types, structure infographics, and provide guidance on effective visual communication. (For using visuals to understand and remember a concept — mind maps and concept maps as a study technique — see Lesson 5: Visualise it.)
7. Exam Preparation and Practice
GenAI can support exam preparation through the generation of practice questions, the creation of study guides, and the simulation of examination scenarios. (This subsection focuses on exam logistics; for active recall as a study technique that strengthens memory, see Lesson 5: Remember it.)
8. Research Skills Development
For students engaged in research projects, GenAI can assist with methodological planning, literature organisation, and understanding research frameworks—whilst recognising that all research outputs must be based on credible academic sources, not AI-generated content.
Critical Guidelines for Practical Application
When employing any of the strategies outlined above, it is essential to:
- Verify permission: Confirm that GenAI use is permitted in your specific unit and assessment before implementation
- Maintain records: Document all prompts used and outputs generated for transparency and potential acknowledgment requirements
- Critically evaluate: Never accept GenAI output at face value; always verify information against credible academic sources
- Preserve authenticity: Ensure that your submitted work represents your own thinking, analysis, and expression
- Iterate and refine: Use multiple prompt iterations to improve output quality rather than accepting first results
- Protect privacy: Never input personal, confidential, or sensitive information into public GenAI platforms
- Acknowledge use: Follow your unit coordinator's guidelines for acknowledging and citing GenAI assistance
The most effective use of GenAI occurs when students employ it strategically as part of a broader learning ecosystem that includes:
- Engagement with course materials, lectures, and tutorials
- Consultation of academic literature from CQUniversity Library databases
- Discussion with peers, tutors, and teaching staff
- Practice and application of skills through authentic tasks
- Critical reflection on learning processes and outcomes
GenAI represents one tool among many—its value is maximised when integrated thoughtfully within a comprehensive approach to learning.
📚 Further Resources
For additional guidance on using GenAI responsibly in your studies:
- Consult your unit's assessment guidelines and course outline
- Contact CQUniversity Study Support services for personalised assistance
- Visit the CQUniversity Library for research and referencing support
- Speak with your unit coordinator if you have specific questions about GenAI use in your assessments
Lesson 5: Cognitive Learning Strategies
Lesson 4 set out the practical tasks AI can support across your studies. This lesson goes one level deeper, into the learning science behind effective study. GenAI is most useful when it helps you practise learning, not when it replaces it. Cognitive Learning Strategies are structured ways of processing information — breaking it into parts, asking deeper questions, summarising it in your own words, strengthening memory, and representing ideas visually — that turn everyday AI use into deeper understanding and stronger recall.
Use GenAI as a study coach. Do not ask it to produce assessable work for you. Always check your unit profile and assessment instructions before using AI in any assessment-related activity.
The Five-Strategy Study Sequence
The following sequence turns AI use into an active learning process. Each strategy keeps the student responsible for understanding, judgement, and final work.
Break a large topic, reading, or task into manageable parts.
Ask questions that connect, compare, apply, and test ideas.
Write your own summary, then ask AI to identify gaps or unclear points.
Create recall questions, mnemonics, or practice quizzes.
Represent the topic as a table, flowchart, timeline, or concept map.
Prompt Templates
Use one prompt at a time. The goal is to guide your learning process, not to create work for submission.
Combined Workflow Prompt
"I am studying [insert topic]. Guide me through five cognitive learning strategies one step at a time: chunk the topic, ask elaboration questions, help me check my own summary, create a recall activity, and suggest a visual representation. Act as a study coach. Do not write an assignment for me."
Write or think first, then use AI to test, question, organise, and improve your understanding.
Keep a record of prompts and outputs if AI use informs assessment preparation.
Do not copy AI-generated summaries, visuals, or explanations into assessed work unless explicitly permitted and acknowledged.
Do not let AI remove the need to practise recall, explanation, writing, or critical judgement.
Cognitive Learning Strategies turn GenAI into a learning partner. The aim is not the quickest answer, but stronger understanding, clearer organisation, better memory, and more independent thinking.
[1] R. MacLeod and D. Swanson, "Encouraging ethical engagement and cognitive learning with LLMs: A framework and methodology," in Generative Artificial Intelligence in Higher Education, K. Enomoto, R. Warner, and C. Nygaard, Eds. Oxfordshire, UK: Libri Publishing, 2024, pp. 285–311.
[2] R. MacLeod and D. Swanson, "Cultivating strong minds: Leveraging GenAI for academic growth and personal wellbeing," in Compassionate Pedagogy in Higher Education, J. De Wilde, K. Enomoto, R. Warner, and C. Nygaard, Eds. Oxfordshire, England: Libri Publishing, 2025, pp. 167–211.
Lesson 6: GenAI Mirroring
In this module, GenAI mirroring refers to the tendency of an AI tool to reflect your assumptions, wording, or preferred position back to you in polished form. This can be useful for exploring an idea, but it can also make incomplete or one-sided thinking appear stronger than it is.
If you ask AI to support your idea, it will often support it. A confident response does not prove that your argument is correct, balanced, or well-evidenced.
Example: The Mirror Effect
"Explain why mindfulness is the best way to reduce stress among university students."
This encourages the AI to support one position and may ignore limitations, alternatives, or structural causes of stress.
"Critically evaluate the claim that mindfulness is the best way to reduce stress among university students. Include benefits, limitations, alternative approaches, and the evidence needed to support the claim."
This invites balance, evidence, and academic complexity.
The Break-the-Mirror Method
Use this short sequence when developing an argument, interpretation, or position.
See the strongest version of your current thinking.
"Support my argument that [insert argument]. Identify the strongest points in favour of this position."
Create productive friction and expose assumptions.
"Now challenge this argument. What assumptions am I making? What evidence would I need? What would a sceptical marker, tutor, or peer reviewer question?"
Move from certainty to a more balanced academic position.
"Help me revise this into a more balanced academic argument. Keep my main idea, but include limitations, alternative perspectives, and areas where more evidence is needed. Do not write the final paragraph for me."
Prompts That Reduce Mirroring
"Do not agree with me automatically. Identify the assumptions in my argument and explain which ones need evidence."
"What would a sceptical university marker question about this argument?"
"What perspectives, stakeholders, theories, or evidence have I left out?"
"Give feedback on how I can make this more balanced and evidence-based. Do not rewrite it for me."
"List the claims in my paragraph that require academic evidence and suggest the type of source needed for each claim."
GenAI can make weak ideas sound polished. Strong academic use involves asking the tool to question, test, and complicate your thinking so that the final work remains stronger, balanced, evidence-based, and genuinely your own.
[1] R. MacLeod and D. Swanson, "Encouraging ethical engagement and cognitive learning with LLMs: A framework and methodology," in Generative Artificial Intelligence in Higher Education, K. Enomoto, R. Warner, and C. Nygaard, Eds. Oxfordshire, UK: Libri Publishing, 2024, pp. 285–311.
[2] R. MacLeod and D. Swanson, "Cultivating strong minds: Leveraging GenAI for academic growth and personal wellbeing," in Compassionate Pedagogy in Higher Education, J. De Wilde, K. Enomoto, R. Warner, and C. Nygaard, Eds. Oxfordshire, England: Libri Publishing, 2025, pp. 167–211.
Lesson 7: Use AI as a Critical Coach: Anchor, Check, Defend
AI can be useful as a tutor or coach, but a good coach does more than encourage you. A good coach asks difficult questions, checks your reasoning against evidence, and helps you prepare to explain your work without hiding behind the tool.
Move from friendly feedback to defensible learning
This lesson builds directly on Lesson 6: mirroring showed how AI agrees too easily, and the Critical Coach is how you push back. Use the three-part model below whenever an AI response sounds convincing but needs to be tested. It puts the SAGE cycle into practice — Anchor and Check operationalise the Evaluate and AI Critic steps, and Defend is the SAGE Defend step itself.
Do not only ask whether an AI answer is good. Ask what it is good against. A response should be checked against a rubric, standard, guideline, paper, policy, code, dataset, textbook, professional principle, or verified source.
1. Coach: Use AI to practise, not to outsource
AI can help you learn by asking questions, giving feedback, explaining difficult points, testing your understanding, or helping you prepare for a discussion. The safest pattern is to think first, draft first, or attempt the task first, then use AI to coach your next improvement.
"Here is my draft answer. Do not rewrite it. Ask me five questions that will help me improve the reasoning, evidence, and structure."
"Write the final answer for me and make it sound academic."
2. Anchor: Force the answer outside the conversation
An anchor is an external reference point that helps test whether an AI response is accurate, relevant, and appropriate. The anchor may be provided by your unit, your discipline, a professional body, a standard, a published paper, or your assessment instructions.
Interactive Anchor Selector
Choose an anchor type to see how a student might ask AI to check an answer without simply accepting the first response.
Use this when the task sits within a recognised discipline or profession, such as cybersecurity, nursing, accounting, education, engineering, social work, design, or construction.
Check this answer against [insert standard/framework/guideline, e.g., NIST CSF, Australian nursing standards, accounting standards, building code, design principle]. Identify where it aligns, where it is incomplete, and which claims I must verify in the original standard. Do not assume the answer is correct.
Early in a unit, you may not know which standard, guideline, theory, code, framework, or key source is most relevant. You can ask AI to suggest possible anchors, but you must still verify them through the unit materials, lecturer, library databases, or official websites.
"I am studying [topic] in [discipline]. What official standards, guidelines, theories, codes, frameworks, or key sources are commonly used to check work in this area? For each one, explain why it may be relevant and how I should verify that it applies to my task."
3. Check: Ask for critique, not agreement
Once you have an anchor, use AI to check the response against that anchor. You can also use a fresh chat, another AI tool, or a stricter role to reduce the chance that the tool simply continues agreeing with the earlier conversation.
🧭 Ask a stricter question
Turn a claim into a question before asking AI to evaluate it. This reduces the chance that the tool simply supports the position you already stated.
"Rewrite my claim as a fair academic question. Then evaluate both sides, identify assumptions, and state what evidence would be needed."
🔁 Start a fresh critique
Paste the AI response into a new chat and ask for critique. A fresh context can reduce conversational momentum from the first chat.
"Do not assume this AI response is correct. Identify unsupported claims, missing evidence, contradictions, and what I should verify against an external source."
👥 Get a second opinion
A different AI tool or a different critical role can reveal gaps, but it is not proof. Treat the second response as another prompt for verification.
"Act as a sceptical marker, reviewer, professional assessor, or subject expert. Critique this answer against [anchor] and list revision priorities."
🧩 Surface criteria first
Before asking for a final answer, ask the tool to identify the criteria, assumptions, evidence, and limitations that should guide the answer.
"Before answering, list the criteria, assumptions, evidence needed, and limitations. Then give a concise answer and tell me what I must verify."
4. Defend: Prepare to explain and adapt
A good AI coach should prepare you to demonstrate your own learning. Your lecturer may ask you to explain your reasoning, modify the work, complete a related task, walk through a process, present your choices, or answer questions under assessment or assurance conditions. This is the same habit you will apply as the Transfer Test in Lessons 10 and 11: if you cannot reproduce or justify the work without the tool, you have not finished learning it.
Defend prompt
Ask me five questions to test whether I can defend this answer. Include questions about evidence, assumptions, limitations, alternative explanations, and how I would respond if the task requirements changed. Ask one question at a time and wait for my answer.
Apply the model across disciplines
- Anchor: What external source, standard, evidence, or criterion should this answer be checked against?
- Check: What assumptions, unsupported claims, gaps, or contradictions does the AI response contain?
- Defend: Can you explain, adapt, and justify the work if your lecturer asks you to demonstrate your understanding?
Activity: Take an AI answer on a verification ride
Choose one AI response you have received. It may be a paragraph, slide outline, design idea, code explanation, data interpretation, policy draft, or study answer. Then:
- Identify the most relevant anchor: rubric, standard, guideline, reading, dataset, code, policy, or professional principle.
- Ask AI to check the response against that anchor.
- Use a fresh chat or another AI tool to critique the first response.
- Open the anchor yourself and verify the most important claim.
- Ask AI to question you until you can defend the answer in your own words.
Lesson 8: Finding Academic Sources
This lesson demonstrates how to use AI tools to help you find 3–5 credible sources for your assignment and work with them effectively. AI assists with planning and understanding—it does not write your work or invent references.
By the end of this lesson, you will be able to:
- Turn your assignment topic into effective search keywords
- Find and verify real academic sources using library databases
- Use AI to summarise complex articles and check your understanding
- Find additional sources based on one good reference you already have
- Acknowledge AI assistance appropriately in your work
Why Real Sources Matter
Your assignment requires evidence from credible academic sources—not AI-generated content. AI tools like ChatGPT cannot access current research databases and frequently invent fake references that look real but do not exist. Therefore, you must always verify sources independently.
✓ Credible Sources Include:
- Peer-reviewed journal articles – Found via your University Library databases or Google Scholar
- Academic books and book chapters – Published by university presses or established academic publishers
- Government and major NGO reports – From organisations like WHO, OECD, Australian Bureau of Statistics
- Professional body publications – Standards, guidelines, and reports from ACM, IEEE, professional associations
✗ Weak Sources to Avoid:
- Websites without named authors or publication dates
- Wikipedia (use it for background only, not as a cited source)
- Commercial websites, blogs, or opinion pieces without peer review
- Any reference that cannot be found in your library or Google Scholar
The Five Ways AI Can Help (and Can't Help)
Turn your assignment topic into focused search terms for library databases and Google Scholar.
Get a plain-language summary of a difficult article you have already found and downloaded.
Compare your notes with AI summaries to identify gaps or misunderstandings.
Use titles and keywords from one good source to discover similar articles.
Check whether your written argument aligns with what sources actually say.
AI invents fake references. Never cite a source you have not personally found and verified.
Submitting AI-generated text as your work constitutes academic misconduct.
You must evaluate sources, synthesise information, and form your own conclusions.
AI tools cannot search library databases or access paywalled journal articles.
You remain responsible for checking that every source is credible and relevant.
Step-by-Step Workflow: Finding Your 3–5 Sources
This workflow takes approximately 30–45 minutes and results in a verified list of credible sources for your assignment.
Practical Example: Complete Workflow
Scenario: Nursing Student Assignment
Assignment question: "Discuss the impact of nurse-to-patient ratios on patient safety outcomes in Australian hospitals."
Step 1: Keywords (AI-generated):
- nurse staffing levels
- patient safety outcomes
- nurse-patient ratios
- hospital staffing Australia
- adverse events nursing
Step 2: Sources found (via Library):
- Journal article from Australian Health Review (2022)
- Government report from Australian Institute of Health and Welfare (2023)
- Peer-reviewed article from Journal of Nursing Management (2021)
Step 3: Verified: All three appear in CQUniversity Library with DOIs, downloadable PDFs, and clear author information. ✓
Step 4: Summarised: Used AI to create plain-language summaries of the abstracts, then compared with my own reading notes to check understanding.
Step 5: Found additional source: The 2022 article cited a WHO report on global nursing workforce—found and added that as a 4th source.
Source Tracking Template
Use this table to organise your sources as you find them. Copy it into a Word document or spreadsheet.
| Search Keywords Used | Full Reference | Found Via | How It Supports My Argument | Verified? (✓ or ✗) |
|---|---|---|---|---|
| nurse staffing levels, patient safety | Smith, J. (2022). Staffing impacts on outcomes. Australian Health Review, 46(3), 245-260. | CQU Library | Provides statistical evidence linking ratios to falls and infections | ✓ Found with DOI |
| [your keywords] | [full citation] | [Library/Scholar] | [definition/example/data/counter-view] | [✓ or ✗] |
| Add 2–4 more rows as needed | ||||
Quality Check: Is This Source Good Enough?
Before including any source in your assignment, answer these four questions:
- Who wrote it? Named researchers, government agencies, or professional bodies = good. Anonymous or unclear authorship = weak.
- When was it published? Last 5 years = ideal for most topics. Older sources acceptable for foundational concepts or historical context.
- Where was it published? Peer-reviewed journals, academic presses, government sites = credible. Personal blogs, commercial sites = weak.
- Can you access the full text? Must be downloadable from Library or available via DOI. If you cannot access it, you cannot verify it = do not cite it.
Using AI to Verify Your Own Writing
Once you have written a draft paragraph, AI can help you check whether your interpretation of sources is accurate.
Academic Integrity and AI Acknowledgement
- Never submit AI-written text as your own work – This includes sentences, paragraphs, or entire sections
- Never cite a reference you have not personally verified – AI invents fake references that constitute academic misconduct
- Never paste confidential or personal information into public AI tools – Use university-endorsed tools where available
- Always acknowledge AI assistance – Even if you only used it for brainstorming keywords
How to Acknowledge AI Use
If you used AI as described in this lesson, include this statement in your assignment (adapt as needed):
"I used [name of AI tool, e.g., ChatGPT] on [date] to generate search keywords for library databases and to create plain-language summaries of article abstracts to check my understanding. I searched for all sources independently using CQUniversity Library and Google Scholar. I verified that every cited reference is real and accessible. All analysis, synthesis, and written content in this assignment represents my own work. No AI-generated text has been submitted as my original writing."
Quick Reference: Safe vs Unsafe AI Use
| ✓ Safe and Permitted | ✗ Unsafe and Prohibited |
|---|---|
| Generating search keywords from your topic | Asking AI to "find references" or "give sources to cite" |
| Summarising abstracts in plain language | Copying AI summaries into your assignment as your analysis |
| Checking your notes against source content | Asking AI to write your literature review or paragraphs |
| Identifying gaps in your understanding | Submitting AI-generated text as your own work |
| Generating additional search terms from good sources | Citing references that AI mentioned without verifying them yourself |
Self-Check Before Moving to the Quiz
Answer these three questions to confirm your understanding:
- Where should you search for academic sources? (Name two places)
- Give one example of safe AI use and one example of unsafe AI use for finding sources.
- What must you do before citing any source in your work?
Show sample answers
1. Where to search: CQUniversity Library databases (primary) and Google Scholar (verification)
2. Safe vs unsafe: Safe = using AI to generate search keywords. Unsafe = asking AI to provide references to cite without verifying them in the library.
3. Before citing: Verify the source exists in your library or Google Scholar, access the full text, read it yourself, and confirm it is relevant and credible.
AI is a research assistant that helps you think and plan—not a replacement for your own searching, reading, and analysis. Every source you cite must be real, verified, and personally evaluated. Your assignment represents your understanding and your voice.
Lesson 9: AI in Group Projects
Group assignments introduce unique challenges when AI tools are available. This lesson addresses how to use AI fairly and transparently within teams, ensure equitable contributions, and maintain both individual learning and collective academic integrity.
Common problems that arise:
- One team member uses AI extensively while others do not, creating unequal effort
- Unclear who contributed what when AI is involved
- Disagreement about whether or how much AI should be used
- One person does minimal work but benefits from AI-assisted contributions by others
- Uncertainty about how to acknowledge AI use in collaborative work
- Risk that some team members do not develop skills because AI did the work
Core Principle: Transparency and Fairness
Successful group work with AI requires agreement, documentation, and honesty. Every team member must understand what AI tools are being used, how they are being used, and who is responsible for verifying and refining AI outputs.
The Three Rules of AI in Group Work
- Establish a team AI agreement before you start – Everyone must agree on acceptable AI use
- Track and document all AI contributions – Know who used AI, when, and for what purpose
- Ensure every team member learns and contributes – AI cannot replace individual skill development
Step 1: Create Your Team AI Agreement (First Meeting)
Before any work begins, your team should complete a short AI use agreement. This prevents conflicts and ensures everyone has the same expectations.
Step 2: Fair Task Division When AI is Available
AI should not create unequal workloads. If one person uses AI to complete their section in 30 minutes while another spends 5 hours without AI, this creates inequity and resentment.
Step 3: Tracking Individual Contributions
When AI is involved, it becomes harder to see who did what. Transparent tracking protects everyone and provides evidence of individual effort.
Step 4: Handling Disagreements About AI Use
Not everyone in your team may feel comfortable with AI, and that is legitimate. Disagreements must be resolved respectfully.
"Can you explain why you're uncomfortable with using AI for this task? I want to understand your perspective."
"What if we use AI only for brainstorming keywords, but write all content ourselves?"
"If you prefer not to use AI for your section, that's fine. Let's adjust task division so workload stays fair."
"We have different views on AI use. Can we discuss this with you to clarify what's acceptable?"
"Everyone uses AI now, you're being old-fashioned."
"We're using AI whether you like it or not."
Using AI for your section without telling the team violates trust and your agreement.
"You don't want to use AI? Fine, you do all the hard work then."
Step 5: Collective AI Acknowledgement
Your team must include a single, clear acknowledgement statement that describes AI use across the entire project.
Template: Group AI Acknowledgement Statement
"This group project was completed by [list all team member names]. Our team used the following AI tools: [ChatGPT / Microsoft Copilot / other] on [dates] for the following purposes: [generating search keywords for literature review / creating project timeline / checking grammar]. All AI outputs were reviewed, verified, and refined by team members before inclusion. AI was not used to write final content, generate references, or replace individual analysis and critical thinking. Each team member contributed to research, writing, and review. A detailed contribution log is available upon request. Final submitted work represents our collective understanding and original thinking."
Alternative if AI was not used:
"This group project was completed by [list all team member names]. Our team did not use generative AI tools for any aspect of this work. All research, writing, analysis, and formatting were completed by team members without AI assistance."
Common Group Work + AI Scenarios
Ensuring Everyone Learns (Not Just Completes Tasks)
The purpose of group work is to develop skills, not just produce a final document. AI should not prevent any team member from learning essential competencies.
Questions to Ask as a Team
- Can everyone explain our project's main argument? If one person used AI heavily, do they still understand the content?
- Could everyone complete their section without AI if needed? Have you learned the skills, or just managed the tool?
- Would we get a good mark if AI disappeared tomorrow? Or are we dependent on it?
- Can everyone defend the work in a presentation or exam? Understanding matters more than completion.
Fair AI use in group work requires clear agreements, transparent tracking, balanced task division, and collective accountability. AI should enhance your team's work without creating inequality, reducing learning, or obscuring individual contributions. When used responsibly, AI can improve coordination and efficiency—but it never replaces the responsibility to learn, contribute, and maintain integrity.
Lesson 10: AI with Your Own Materials (NotebookLM, Codex, Claude Code)
Some AI tools can work directly with the materials you provide. These may include readings, lecture notes, slides, PDFs, textbook chapters, assessment instructions, datasets, drafts, code files, spreadsheets, or connected cloud documents. Examples include NotebookLM, ChatGPT or Claude with uploaded files, Gemini connected to cloud documents, and coding tools such as Codex or Claude Code that can work with files in a project.
This is different from asking a general chatbot a broad question. When AI works with your own materials, it can help you locate evidence, compare sources, test your understanding, inspect code, and identify gaps in your knowledge. However, it can still misunderstand, oversimplify, miss important details, or make your materials appear more complete than they really are.
Use AI to engage more carefully with your own materials. Do not use it to avoid reading, practising, testing, verifying, or thinking.
Learning Objectives
- Recognise when an AI tool is working from your own files, notes, code, sources, or project materials.
- Use AI to question, compare, and verify materials rather than passively accept summaries.
- Trace AI responses back to the original source, file, page, paragraph, slide, dataset, or code section.
- Identify when an AI response may be incomplete, oversimplified, or misleading.
- Use AI to prepare for assessment tasks that require explanation, adaptation, or defence of your work.
What Changes When AI Can See Your Files?
When an AI tool can access your files, it may appear more reliable because it is using material you uploaded or connected. This can be useful, but it does not make the tool automatically correct. The tool can only work with what it can see. If the file set is incomplete, outdated, duplicated, or one-sided, the AI response may also be incomplete or one-sided.
Four Good Ways to Use File-Connected AI
Click each card to see the prompt template. In all four cases, the goal is to keep you active — thinking, checking, and practising, not passively reading a polished AI summary.
The Transfer Test: Can You Still Do It Yourself?
Sometimes students use tools such as Codex, Claude Code, Copilot, ChatGPT, Gemini, or other systems to produce or modify work. A realistic learning rule is needed: if AI helped produce part of your work, you still need to understand the process and be able to apply the skill again.
If AI helped you produce an output, you should be able to explain it, check it, change it, and reproduce the same type of work under different conditions. You will return to this test in Lesson 11, where AI can produce the artefact itself.
Your lecturer may test this transferable learning through an exam, oral discussion, practical demonstration, code walkthrough, presentation, modified requirement, or another assurance activity. The point is not only whether the final artefact looks correct. The point is whether you developed the skill the assessment was designed to measure.
Watch Out: The False Completeness Problem
A file-connected AI may produce an answer that sounds complete, but that does not mean it has covered the whole topic. It may only reflect the materials you uploaded. If important sources, files, data, requirements, or perspectives are missing, the answer may also be missing important information.
"Based on the uploaded materials, identify what this source set does not cover. What additional evidence, perspectives, data, requirements, tests, or readings would I need before making a strong claim or submitting this work?"
Quick Check: Read, Trace, Test
- Read: Have I engaged with the original material myself?
- Trace: Can I trace the AI's answer back to the exact source, file, or code location?
- Test: Can I answer questions or perform a related task without simply repeating the AI output?
Activity: The Source Trace Challenge
Choose one reading, slide deck, dataset, draft, code segment, spreadsheet, or technical file from your unit.
- Upload or connect the file to an AI tool that can work with your material.
- Ask the tool to identify three important claims, ideas, findings, steps, or decisions from the file.
- Ask it to provide the exact source location for each one.
- Open the original file and check whether each point is accurately represented.
- Write a short reflection: Which AI response was accurate, which response needed correction, and what did I still need to check myself?
If you used a file-connected tool (such as NotebookLM, or ChatGPT or Claude with your uploads) to study or to prepare assessable work, acknowledge it as you would any AI assistance: name the tool, the date, what you uploaded, and how it helped. File-connected tools blur the line between "my notes" and "AI output," so a clear record of what you provided and what the tool produced is your best protection. Always follow your unit's specific acknowledgement instructions.
AI tools that work with your own materials can help you study more effectively, but only when you remain actively involved. Use them to locate, question, compare, trace, test, and practise your understanding. Do not use them to replace reading, practising, verifying, or thinking.
Lesson 11: AI Inside Your Apps (e.g., PowerPoint, Spreadsheets, Code)
AI tools are no longer limited to separate chat windows. Many can now work inside the applications students use to produce academic work. This includes documents, slides, spreadsheets, forms, browsers, coding environments, notebooks, cloud drives, design tools, data-analysis platforms, and other digital workspaces.
These tools can draft, edit, format, generate slides, write code, fill tables, analyse data, create workflows, or make changes directly inside files. They can be useful, but they can also create academic integrity risks because they may produce part of the work that is submitted for assessment.
What did the AI do, and can you explain, verify, adapt, and defend the result?
Learning Objectives
- Recognise when AI is advising you and when it is producing work for you.
- Use embedded and agentic AI tools responsibly across different disciplines.
- Apply the Transfer Test before submitting AI-assisted work.
- Check AI-generated outputs for accuracy, appropriateness, completeness, and permitted use.
- Understand why lecturers may use exams, oral discussions, demonstrations, walkthroughs, or other assurance activities to test transferable learning.
Advising Versus Doing
This is the single most important idea in the lesson. Not all AI use carries the same risk: lower-risk use involves AI helping you think; higher-risk use occurs when AI produces the artefact itself. Most uses sit somewhere on a spectrum between the two.
Suggesting questions, helping you plan, explaining a concept after you have tried, identifying gaps, or giving feedback without rewriting.
Generating final wording, creating slides, filling forms, writing code, making formulas, producing analysis, or changing the submitted file directly.
Higher-risk use is not automatically prohibited in every unit. Some assessments may allow AI assistance. Others may restrict it. The important point is that the more directly AI contributes to the final submitted artefact, the stronger your responsibility becomes.
The Transfer Test, Applied to Production
You met the Transfer Test in Lesson 10: if AI helped produce part of your work, you should be able to explain it, check it, change it, and reproduce the same kind of work under different conditions. It matters most here, where the tool can generate the artefact itself.
This is because your lecturer may assess not only the final artefact, but also the learning behind it — through an exam, an oral discussion, a modified requirement, a walkthrough of your code, spreadsheet, design, report, or presentation, or by asking you to justify your assumptions, sources, calculations, and methods. This is exactly what the SAGE Defend step is designed to verify; see the Defend Tool for how educators design these checkpoints.
The SAGE Assessment Policy spells out what this means for you — including what happens if you cannot fully defend AI-assisted work that you submit.
Examples Across Disciplines
| AI-assisted output | What you must still be able to do |
|---|---|
| Presentation or slide deck | Explain the argument, evidence, slide order, visual choices, and why each point is included. |
| Code, application, or notebook | Explain the logic, variables, functions, dependencies, tests, errors, edge cases, and possible improvements. |
| Spreadsheet or data analysis | Explain the data source, formulas, assumptions, calculations, method, limitations, and interpretation. |
| Report or written response | Explain the claims, sources, reasoning, limitations, and how the structure supports the task. |
| Form, template, policy, design brief, clinical scenario, or business model | Explain why each entry is accurate, appropriate, relevant, and supported by the task requirements. |
A Responsible Workflow for AI Inside Apps
Special Notes by Output Type
Open the note that matches what you are producing.
Before You Submit: Explain, Adapt, Reproduce
- Explain: Can you explain what the output means and how it was produced?
- Adapt: Can you change it if the task, data, audience, or requirements change?
- Reproduce: Can you complete a similar task under exam, interview, oral, practical, demonstration, or assurance conditions?
Activity: The Transfer Test Challenge
Choose one AI-assisted output. This may be a paragraph, slide, table, formula, code segment, design, workflow, form, data analysis, or other artefact.
- Identify which parts were created, edited, suggested, or changed by AI.
- Explain the purpose of each major part in your own words.
- Identify one possible error, assumption, limitation, or weakness.
- Modify the output for a slightly different requirement.
- Write a short reflection: What did I learn from using AI here, and what would I still need to practise without the tool?
Responsible use of AI inside apps means more than producing a polished output. You must be able to explain, verify, adapt, reproduce, and acknowledge AI-assisted work where required.
Final Quiz: Check Your Understanding
This quiz has been designed to assess your knowledge of the key concepts from this module. Select your answers and receive immediate feedback.
This quiz consists of twelve multiple-choice questions followed by three short-answer reflection questions. Select the best answer for each multiple-choice question, then click "Submit Quiz" to receive your results and feedback.
Your Results
Reflection Questions
The following questions require written responses. Consider them carefully and use them to reflect on your learning.
Task: Rewrite this to be a more effective role prompt. Give the AI a specific persona and provide clear context and instructions.
"Act as a project management tutor helping me structure my reflective writing. I need to write a 500-word reflection on a challenging teamwork experience where communication breakdown led to missed deadlines in my group project. Help me create an outline that follows Gibbs' Reflective Cycle (Description, Feelings, Evaluation, Analysis, Conclusion, Action Plan). For each section, ask me 2-3 probing questions that will help me develop my critical reflection, but do not write the reflection for me."
1. Verify the reference exists: Search for the article using the CQUniversity Library databases, Google Scholar, or the journal's website. AI frequently invents "hallucinated references" that appear legitimate but are completely fabricated.
2. Evaluate the source quality: If the reference exists, assess whether the journal is peer-reviewed, whether the author is credible in the field, and whether the content is relevant and current for your purposes. Never cite a source without reading and evaluating it yourself.
Why this matters: Academic integrity requires that all cited sources are accurate, credible, and actually consulted. Using fabricated or unverified sources undermines your work and violates academic standards.
Strategy 1: Use the Anchor, Check, Defend model. Before relying on an AI answer, I will check it against a relevant anchor such as a rubric, guideline, standard, research paper, dataset, or unit reading. I will then identify gaps or assumptions and practise explaining the answer in my own words.
Strategy 2: Use AI for questioning, not simply answering. Instead of asking AI to write my essay or solve my problem, I will use it to ask probing questions, challenge my assumptions, and test whether I can adapt the answer if the task requirements change.
You have completed the CQUniversity Generative AI Literacy & Numeracy Module. Remember: GenAI is a powerful tool when used responsibly. Always prioritise your learning, maintain academic integrity, and use these technologies to enhance—not replace—your critical thinking and creativity.
Module Wrap-Up & Next Steps
You have taken an important step in developing your digital literacy skills. This capability will serve you well throughout your studies and into your future career.
Your Three Key Takeaways
As you move forward, keep these three core principles in mind:
Your Unit Profile is the single source of truth. Never assume AI is permitted—always verify.
AI is a "co-pilot," not an auto-pilot. You are responsible for every word you submit. Critically evaluate and fact-check.
Use AI to support your thinking, not replace it. If you can't explain it in your own words, you haven't learned it yet.
National Resources & Guidelines
For authoritative guidance on AI in higher education:
National Guidelines & Resources TEQSA Assessment Design:
AI in Assessments TEQSA Academic Integrity:
Guidance for Institutions
CQUniversity-Specific Resources
For the most current and detailed CQUniversity policies:
Official GenAI Referencing Guide Academic Learning Centre (ALC) CQUniversity Academic Integrity Policy
References
[1] M. Elkhodr and E. Gide, "The SAGE framework for developing critical thinking and responsible generative AI use in cybersecurity education," Discover Education, vol. 4, article 517, Nov. 2025, doi: 10.1007/s44217-025-00935-3.
This module is informed by the SAGE framework's published research and the authors' broader work on generative AI in higher education. For the full evidence base and reading list, see the Evidence & Impact page. The contributed lessons (5 and 7) also list their own sources within each lesson.
How to cite this module
IEEE
M. Elkhodr and E. Gide, "Generative AI Literacy & Numeracy Module," Central Queensland University, Rockhampton, Australia, Version 2.0, Jun. 2026. [Online]. Available: https://sage-framework.com/genai-101.html. DOI: 10.5281/zenodo.18295619. [Accessed: 22-Jun-2026].
Harvard
Elkhodr, M. and Gide, E., 2026. Generative AI Literacy & Numeracy Module. Version 2.0. Rockhampton: Central Queensland University. DOI: 10.5281/zenodo.18295619. Available at: https://sage-framework.com/genai-101.html (Accessed 22 June 2026).
If citing a specific lesson, append the section title and version (e.g., "Lesson 3: Responsible Use, v3.0").
About This Module
This module has been developed as a companion resource to the Structured AI-Guided Education (SAGE) framework [1], supporting the development of AI literacy among first-year students.
Development Team: Dr Mahmoud Elkhodr (Senior Lecturer, Cybersecurity & AI Education) and Professor Ergun Gide (School of Engineering and Technology).
Version 2.0 — Lesson contribution acknowledgement: Two lesson-level contributions (Cognitive Learning Strategies and GenAI Mirroring) were provided by CQUniversity's Academic Learning Centre and adapted for integration within the SAGE GenAI-101 module. These lessons draw on MacLeod and Swanson's work on ethical engagement, cognitive learning with LLMs, academic growth, and student wellbeing. The overall SAGE Framework, GenAI-101 module structure, responsible-use model, and integration logic remain part of the SAGE work developed by Elkhodr and Gide.