AI For Staff
🤖 AI Guides @ JCU
General Guidance on AI at JCU
Artificial Intelligence (AI) tools such as Microsoft Copilot, ChatGPT, and other generative or analytical systems are now integral to many aspects of academic, administrative, and professional work at JCU. This guidance applies to all JCU staff, academic, professional, and technical, and outlines the University’s expectations for responsible, ethical, and effective use of AI.
1. Purpose and Principles
JCU encourages the thoughtful and transparent use of AI to enhance productivity, creativity, teaching, research, and service delivery. Staff are expected to:
- Use AI as a support tool, not as a substitute for professional or academic judgement.
- Maintain accuracy, accountability, and confidentiality in all AI-assisted work.
- Ensure AI use aligns with JCU’s values, relevant policies, and applicable legislation (e.g. intellectual property, privacy, and data security).
2. Responsible and Ethical Use
- Transparency: Declare when AI tools have been used to generate text, analysis, or media, particularly in academic, policy, or communication outputs.
- Critical Review: Always check AI-generated information for factual accuracy, bias, and ethical implications before use or dissemination.
- Data Privacy: Never input confidential, personal, or proprietary JCU information into external AI systems that are not covered by institutional agreements. Currently, this can only be done via Microsoft Copilot by logging into the AI tool with your JCU credentials.
- Sustainability and Equity: Recognise that AI systems have environmental costs and accessibility implications; use them judiciously and support colleagues or students who may choose not to use AI.
3. Prescribed and Approved Tools and Systems
- JCU clearly differentiates between:
- Prescribed tools (approved and supported by JCU)
- Public / non-JCU tools (not approved and unsupported by JCU).
- Microsoft Copilot and AI-enabled Microsoft 365 apps are JCU's main prescribed AI tools, secure for JCU use under institutional licensing.
- Prescribed AI tools are the recommended tools to use at JCU and come with a level of data protection. Tools and use cases are listed on the GenAI Tools at JCU page.
- Public/non-JCU AI tools (e.g., ChatGPT, Claude, Gemini) may be used for ideation, drafting, or exploration, provided no teaching materials and sensitive data are shared, and results are critically evaluated. Check the GenAI Tools at JCU page for smart use of non-JCU AI tools.
- Do not upload teaching materials (slides/recordings/notes), assessment items/solutions, third‑party licensed teaching content, or student work to non‑JCU AI tools.
4. AI in Teaching, Research, and Operations
- Academics should model responsible AI use for students integrating AI where it adds learning value and offering clear guidance on acceptable use in assessments and course design.
- Professional staff may use AI to streamline administrative tasks, generate content drafts, or analyse information, while maintaining confidentiality and data integrity.
- Researchers must comply with JCU’s research ethics and data management policies when using AI for data analysis, writing, or image generation.
AI Community of Practice (AI CoP), as part of AI@JCU, aims to serve as a platform to share experiences, use cases, technical advances and insights in the rapidly evolving Artificial Intelligence space.
Join the AI CoP today!
Prompt Engineering and Exemplars
The following prompts are explicitly for Copilot (JCU) or the in-built AI in LearnJCU. If using public / non-JCU AI tools, ensure you rewrite as abstract prompts that do not require pasting of slides or teaching material.
🦺 Safer workflows (“assistance” without the risk)
✓ Work inside JCU prescribed tools first: Draft with Copilot (signed-in) or LearnJCU AI, then optionally polish language in a public tool without pasting internal content.
If working in public / non-JCU tools:
✓ Abstract the prompt: Describe the task and context, not the document (e.g., “Create a 5-step tutorial outline on [topic] for 2nd-year undergrads”).
✓ Do not upload teaching materials: Slides/recordings/notes, assessment items/solutions, third‑party licensed teaching content, or student work to non‑JCU AI tools.
Why prompt engineering matters
The strength of any AI‑generated response begins with the way we speak to the system. A well‑crafted prompt is like giving directions to a visitor in a new city: the more precise and purposeful the instructions, the more likely they are to arrive at the destination you intended. In teaching and curriculum design, prompts shape whether AI delivers scattered ideas or structured, classroom‑ready outputs.
Core principles of effective prompting
Master these basics for immediate improvements.
Lead with intent: Begin every prompt with a clear statement of what you want—summary, lesson plan, or discussion example.
Add meaningful context: Provide the “why” and “who.” Mention student level, subject area, and desired complexity.
Guide the format: Tell the AI the output shape—bullet points, tables, short summaries, or narratives.
Break down the task: Divide complex requests into smaller, ordered steps—treat the prompt like a mini workflow.
Check for clarity and flow: Before sending, ask: would someone outside my discipline understand what I’m asking the AI to do?
Tip for academics
Imagine you are giving instructions to a new teaching staff who has never seen your course before. That is the level of clarity and specificity an AI tool thrives on.
Prompt examples (adapt to your subject)
Module Outlines
“Create a 4 week module plan on renewable energy transitions. Include weekly topics, one key activity per week, and a suggested formative assessment. Present as a simple table.”
Lecture Summaries
“Summarise the essential points from my Week 2 lecture slides on data visualisation. Limit to 250 words and include two reflective questions for students.”
Contextual Case Studies
“Design a case study showing a dilemma faced by a school implementing AI based plagiarism detection. Include roles for a teacher, student, and administrator, and finish with discussion prompts.”
Teaching with Prompt Engineering
Why use AI for teaching design
AI can be a creative teaching partner if we guide it thoughtfully. Effective prompts can help generate new examples, simplify complex ideas, or adapt materials for students at different levels. Think of AI as a collaborator that reflects the clarity of your instructions—vague prompts lead to generic outputs, while precise prompts can spark engaging learning moments.
Practical prompting strategies for teaching
Set the classroom scene
Indicate the student level (e.g., “first-year undergraduate”) and the learning environment (e.g., “online forum” or “lab”).Request varied perspectives
Ask for analogies, step-by-step reasoning, or comparative explanations.Encourage active learning
Include instructions to create prompts, questions, or short exercises that get students thinking.Model inclusivity
Prompt AI to consider diverse backgrounds and provide accessible explanations free of jargon.
Prompt examples (adapt to your subject)
Simplifying a concept
“Explain photosynthesis to first year science students using a cooking metaphor, then provide a one sentence takeaway they can remember.”
Preparing discussion starters
“Create five thought provoking questions for a class debate on the ethics of using AI in criminal sentencing.”
Multiple explanation levels
“Describe Bayes’ Theorem in three ways: a short analogy for beginners, a clear technical explanation for intermediates, and a practical example for advanced learners.”
Assessment with Prompt Engineering
Why prompt engineering matters for assessment
Assessments designed with AI support can save time in creation while ensuring alignment with learning outcomes. Clear prompts allow AI to generate draft questions, rubrics, or feedback ideas—but human judgment is essential to refine, adapt, and maintain academic integrity.
Approaches to using AI in assessment design
Generate question banks
Ask for variations of questions that target different cognitive levels, from comprehension to evaluation.Draft rubrics and marking guides
Prompt AI to propose criteria, weightings, and performance descriptors aligned to your subject outcomes.Simulate student responses
Have AI attempt your draft questions to identify gaps, potential misinterpretations, or tasks that might be too easy for AI alone.Create feedback templates
Ask AI to produce adaptable, constructive feedback statements for common issues you expect to see.
Tip for academics
Use AI to accelerate the design of assessments, not to replace your professional judgment. The goal is to spark ideas, not outsource responsibility.
Prompt examples (adapt to your subject)
Question generation
“Produce 3 short answer and 2 scenario based questions on coastal erosion management. Ensure each requires critical thinking and not just recall.”
Rubric drafting
“Draft a marking rubric for a 1,500 word reflective essay on sustainable business practices, with criteria for insight, evidence, structure, and style.”
Feedback templates
“Suggest 5 feedback statements I can adapt for students who submit projects with strong analysis but weak visual presentation.”