Table of Contents
- Lesson Overview
- Learning Outcomes
- Before You Begin
- Retrieval Practice
- Opening Scenario
- Core Content
- Learning-Process Walkthrough
- Worked Example
- Guided Learning Activities
- Exercises
- Homework Problems
- Mini-Project
- Self-Check Quiz
- Assignment Connection
- Journal Prompts
- Key Takeaways
- Key Terms
- Looking Ahead
- Sources & Further Reading
Lesson Overview
This is the culminating lesson of Unit 12. You will integrate all the concepts you have learned — learning goals, assistance vs. substitution, productive struggle, scaffolding, feedback, retrieval practice, reconstruction, transfer, metacognition, and cognitive offloading — into a coherent framework: the LEARN framework (author-approved expansion pending). You will also learn how to apply TRACE and CLAIM to verify any AI-generated material you use for learning. You will design a personal AI learning workflow that preserves your intellectual responsibility, establishes clear boundaries for AI use, and includes robust documentation and disclosure practices. Finally, you will evaluate whether AI use actually improved your learning or merely accelerated task completion. This lesson prepares you for Unit 13, where you will apply these principles to research and scholarly practice.
Lesson Learning Outcomes
By the end of this lesson, you will be able to:
- Apply the author-approved LEARN framework (or its operational components) to a learning activity.
- Integrate TRACE and CLAIM into AI-supported learning.
- Design a complete human-governed learning workflow that includes independent effort, bounded AI assistance, verification, retrieval, reconstruction, transfer, and reflection.
- Establish clear assessment, privacy, integrity, verification, and no-AI boundaries.
- Document AI use and learning activities appropriately.
- Evaluate whether AI improved learning rather than merely accelerating completion.
- Identify learning activities that should remain entirely human-led.
- Define human referral and stop conditions for AI use.
- Prepare for responsible AI use in research and scholarly practice (Unit 13).
Before You Begin
- Assigned reading: Chapter 12: Generative AI for Learning (sections on LEARN, workflows, and boundaries).
- Supporting readings: Chapter 7 (TRACE), Chapter 8 (CLAIM), and selected sections of Chapter 13.
- Prerequisite knowledge: You should be comfortable with all concepts from Lessons 12.1–12.3.
- Optional AI tool: You may use any generative AI system, but many activities are designed for reflection and planning. No paid service is required.
- Academic-integrity reminder: This lesson is for learning practice. Do not use AI to generate personal reflections or assessed work.
- Privacy reminder: Do not enter sensitive personal information into AI systems.
- Accessibility reminder: Contact accessibility services if you need accommodations. AI-generated adaptations must be verified.
- Estimated preparation time: 20 minutes to review the reading and previous lessons.
Retrieval Practice: Units 10, 11, and 12
Before we integrate everything, retrieve key concepts from earlier units. Attempt each question from memory before opening the answer.
1. What is the difference between AI-supported learning and AI-substituted learning?
Show answer
Answer: AI-supported learning occurs when AI helps the learner engage in and strengthen the intended learning activity; AI-substituted learning occurs when AI performs the cognitive work that the learner is expected to develop.
2. What is productive struggle and how does it differ from unproductive difficulty?
Show answer
Answer: Productive struggle is sustained effort on an appropriately challenging task that leads to learning. Unproductive difficulty is caused by inaccessible design, lack of prerequisites, or insufficient support.
3. What is the difference between recognition and recall?
Show answer
Answer: Recognition is identifying information when it is presented; recall is producing information without cues. Recall is a stronger indicator of learning.
4. What is cognitive offloading, and when is it undesirable?
Show answer
Answer: Cognitive offloading is transferring cognitive tasks to external tools. It is undesirable when it replaces foundational learning, independent reasoning, or necessary practice.
5. What is the primary purpose of TRACE and CLAIM in learning?
Show answer
Answer: TRACE helps identify errors and uncertainties in AI-generated content; CLAIM provides a systematic process to verify claims using external sources. Both are essential for ensuring the accuracy of material you learn from AI.
Opening Scenario: A Complete AI-Supported Learning Session
A student, Taylor, is preparing for an assignment in a public health course. Taylor uses AI in the following ways:
- Asks AI to summarise the assigned reading.
- Uses the summary to generate an outline for the assignment.
- Asks AI to draft several paragraphs of the analysis.
- Asks AI to generate citations for the arguments.
- Uses AI to assess the draft and suggest improvements.
- Finally, asks AI to write a personal reflection on the learning process.
Taylor then submits the final product, including the reflection, without any verification or documentation of AI use.
Questions for reflection:
- Which of Taylor's uses are potentially useful support? Which are unverified support? Which are substituted learning?
- What integrity violations might be occurring?
- What disclosure is required, if any?
- Which activities should Taylor redesign or stop altogether?
Show guided analysis
Analysis:
- Potentially useful support: Using AI to summarise the reading (if verified) and generating an outline (if Taylor does the actual writing). However, these are only supportive if Taylor engages critically with the AI output.
- Unverified support: The summary, outline, draft, citations, and feedback are all unverified. Taylor does not check any of the AI's claims or sources.
- Substituted learning: Asking AI to draft the analysis, generate citations, and write the personal reflection are clear substitutions. The AI is performing the core analytical, evaluative, and reflective work that Taylor should be developing.
- Integrity violations: Submitting AI-generated analysis and reflection as one's own work is plagiarism/academic misconduct. The lack of disclosure violates course policy if AI use is required to be documented.
- Redesign: Taylor should read the source material themselves, use AI only for clarification or feedback on their own drafts, verify all AI-generated content, and write their own reflection. They should also disclose AI use according to course policy.
Key insight: A complete learning workflow must include independent effort, verification, and documentation. The AI should support, not replace, the learner's cognitive work.
Core Content
1. The LEARN Framework
LEARN is a framework for responsible AI-supported learning. The author-approved expansion is pending, but operationally it addresses the following components, which we will use in this lesson:
Until the exact expansion is confirmed, we use the operational understanding as defined in the course materials: identification of learning goals; independent learner effort; appropriate AI assistance; retrieval, reconstruction, and verification; and reflection on learning, transfer, boundaries, and next steps.
Based on this operational definition, the LEARN framework includes these key elements:
- Learning goal: Define what you are expected to develop.
- Independent effort: Attempt the task without AI first.
- Appropriate AI assistance: Use AI in a bounded, supportive way that preserves your responsibility.
- Retrieval, reconstruction, and verification: Test your recall, rebuild knowledge from memory, and verify all AI-generated content.
- Reflection, transfer, boundaries, and next steps: Reflect on what you learned, apply it to new contexts, set clear boundaries, and plan future actions.
Each element is accompanied by guiding questions, worked examples, and common misuses. The framework is not a guarantee of learning; rather, it is a structured process to increase the likelihood that learning occurs.
We will now examine each component in detail.
2. Integrating TRACE and CLAIM into Learning
Every AI-generated explanation, hint, or piece of feedback you use must be verified. TRACE helps you identify potential errors, omissions, misleading reasoning, and uncertainty. CLAIM provides a systematic process to verify claims using external sources.
When you encounter an AI-generated learning resource, follow this process:
- Identify claims – What statements is the AI making?
- Classify possible errors – Use TRACE: is there a factual error? Reasoning error? Hallucination? Uncertainty?
- Identify evidence requirements – What evidence would be needed to support the claim?
- Evaluate sources – Use CLAIM: locate the source, assess authority, check corroboration.
- Correct or qualify – Revise the explanation based on verification.
- State remaining uncertainty – Be honest about what you still do not know.
This verification process should be part of your learning workflow. Never accept AI-generated content as authoritative without verification.
Example: AI explains the concept of "confirmation bias." You use TRACE to check for errors, then CLAIM to verify the definition against a psychology textbook. You find that the AI's definition is accurate but lacks a key nuance about the role of motivated reasoning. You add that nuance to your own understanding.
3. Designing a Personal AI Learning Workflow
A personal AI learning workflow is a structured sequence of steps you follow when using AI to support your learning. It ensures that you maintain control and responsibility. A complete workflow includes:
- Learning goal – What do you need to achieve?
- Independent attempt – Try the task without AI.
- Bounded AI assistance – If stuck, use AI for a specific, limited purpose (e.g., clarification, hint, feedback).
- Verification – Check the AI's contribution using TRACE and CLAIM.
- Retrieval – Test your recall without AI.
- Reconstruction – Rebuild the knowledge from memory.
- Transfer – Apply it to a new context.
- Reflection – Evaluate what worked, what didn't, and what you still need to learn.
- Documentation – Record your AI use, prompts, outputs, and verification.
- Human referral – Seek human help when needed.
This workflow is flexible; you may not need every step for every activity. But it provides a robust template for responsible AI use.
4. Establishing Learning Boundaries
Boundaries define what is allowed, what requires verification, what requires disclosure, and what is entirely off-limits. A Learning-Boundary Charter helps you make these decisions explicit.
Consider these categories:
- Permitted practice uses – e.g., brainstorming ideas, generating practice questions, clarifying concepts.
- Uses requiring verification – e.g., factual explanations, citations, data.
- Uses requiring disclosure – any AI contribution to assessed work (if permitted).
- Course-dependent uses – follow the specific assignment rules.
- No-AI activities – tasks that must be done entirely by you (e.g., personal reflections, final exams, unaided problem sets).
- Protected information – never share personal, health, or confidential data with AI.
- Human-support needs – when you will consult an instructor, tutor, or other human.
- Stop conditions – signals that you should limit or stop AI use.
Create your own charter and review it regularly.
5. Documentation and Disclosure
Proper documentation helps you track your learning process and ensures transparency. For any AI use in a learning activity, you should record:
- Learning goal
- Initial understanding
- Independent attempt (what you did without AI)
- AI prompt (exact wording)
- Nature of AI assistance (hint, explanation, feedback, etc.)
- AI output
- Verification process and corrections
- Reconstructed answer or explanation (from memory)
- Transfer task and performance
- Reflection on learning
- Disclosure statement (if required)
Documentation is not just for compliance; it is a learning tool that helps you reflect and improve.
6. Evaluating Whether AI Improved Learning
Using AI can save time, but that does not mean it improves learning. To evaluate effectiveness, compare these factors with and without AI use:
- Time to initial output
- Preparation time (reading, planning)
- Verification time
- Correction time
- Independent recall (can you explain it without AI?)
- Reconstruction quality
- Transfer performance
- Confidence calibration
- Total learning value (not just task completion)
If AI accelerates completion but weakens your ability to recall, reconstruct, or transfer, then it has not improved learning — it has merely substituted for it.
7. Human Referral and Stop Conditions
You should consult a human (instructor, tutor, librarian, etc.) when:
- Course expectations are unclear.
- Accessibility support is required.
- The AI response conflicts with course material.
- Sensitive personal circumstances arise.
- Professional judgment is needed.
- Feedback affects formal grading.
- A source cannot be verified.
- Persistent misunderstanding remains.
- AI use may be prohibited.
You should stop or limit AI use when:
- The system begins completing assessed work.
- Assistance exceeds the permitted boundary.
- You have not attempted the task independently.
- The output introduces repeated errors.
- You are copying rather than reconstructing.
- The interaction requests sensitive information.
- You cannot explain the result.
- Confidence increases without demonstrated performance.
- The tool replaces necessary human support.
- Course policy prohibits the use.
8. Transition to Research and Scholarly Practice
In Unit 13, you will apply these principles to research contexts. You will develop research questions, conduct literature searches, synthesise scholarly sources, and adhere to research ethics. The skills of verification, documentation, and responsible AI use are directly transferable. You will also encounter the RIGOUR framework, which builds on LEARN for scholarly practice. This lesson equips you with the foundational habits of mind for that unit.
Learning-Process Walkthrough: Applying LEARN with TRACE and CLAIM
This walkthrough demonstrates a complete learning interaction using the LEARN framework, with TRACE and CLAIM integrated.
- Learning goal: Define and distinguish internal, external, and construct validity; identify threats in a research scenario.
- Independent attempt: The student writes a preliminary definition of each type from memory and applies them to a sample study.
- Bounded AI assistance: The student asks: "Can you provide a concise comparison of internal, external, and construct validity, with an example of a threat to each?"
- AI output: The AI gives a clear comparison with examples.
- Verification (TRACE + CLAIM): The student checks the AI's definitions against a research methods textbook. They find that the internal validity definition is accurate, but the example for construct validity is slightly off. They correct it using the textbook.
- Retrieval: The student closes all sources and writes definitions and examples from memory.
- Reconstruction: The student reconstructs the corrected comparison in a table format.
- Transfer: The student finds a new research article and identifies validity threats in that study.
- Reflection: The student reflects on what they learned and notes that they need more practice with external validity.
- Documentation: The student records the AI prompt, output, corrections, and their own retrieval and transfer attempts.
- Boundary check: This was a practice activity, so no disclosure was required, but the student still documented it.
- Stop condition: The student did not use AI for the retrieval or transfer steps.
Show solution design: Key decisions
- Learning goal: Define and distinguish validity types.
- Independent attempt: Preliminary definitions.
- AI assistance: Comparison and examples (bounded).
- Verification: TRACE and CLAIM against textbook.
- Retrieval: Write from memory.
- Reconstruction: Table format.
- Transfer: Apply to a new article.
- Reflection & documentation: Record and plan next steps.
This design ensures that the learner does the cognitive work and uses AI as a resource, not a substitute.
Worked Example: A Complete LEARN-Based Learning Activity
This worked example demonstrates a full LEARN-based interaction with a synthetic scenario.
A student in a business communication course is learning to write a persuasive memo. The learning goal is to structure a memo that clearly states a problem, proposes a solution, and provides evidence.
| LEARN Component | Student Action | AI Role | Verification / Reflection |
|---|---|---|---|
| Learning goal | Student defines: "I will write a memo that persuades my supervisor to adopt a new project management tool." | — | — |
| Independent attempt | Student drafts a memo structure and content without AI. | — | Student reflects: "My argument is weak; I need better evidence." |
| Bounded AI assistance | Student asks: "What types of evidence are persuasive in a business memo? Give examples." | Assistant: provides types of evidence (data, testimonials, case studies). | Student verifies with a business writing guide (CLAIM). |
| Retrieval | Student writes a new draft from memory, incorporating the evidence types. | — | Student compares to the verified guide. |
| Reconstruction | Student reconstructs the final memo structure in a checklist format. | — | Student checks completeness. |
| Transfer | Student writes a memo for a different scenario (e.g., proposing a new training program). | — | Student evaluates if the structure transfers. |
| Reflection & boundaries | Student reflects: "I learned to structure persuasive memos. I will not use AI to write the memo itself, only for brainstorming evidence." | — | Student documents the AI prompt and output. |
Show analysis
Analysis: This student used AI only for brainstorming evidence types, which is supportive. The core work (drafting, structuring, writing) was done independently. The student verified the AI's suggestions and reflected on the process. The workflow preserved learner responsibility and produced learning evidence (the transfer memo).
Key insight: LEARN is not a linear checklist; it is a cycle of planning, doing, and reflecting.
Guided Learning Activities
Instructions: Choose a learning activity (e.g., learning a concept, solving a problem). Walk through the LEARN components operationally: define the learning goal, attempt independently, use AI for a bounded purpose, verify, retrieve, reconstruct, transfer, reflect, and set boundaries. Document your process.
Show sample walkthrough (for "opportunity cost")
Learning goal: Define opportunity cost and apply it to decision-making.
Independent attempt: Wrote a definition and gave an example from my life.
AI assistance: Asked AI: "Give me a workplace example of opportunity cost."
Verification: Checked against economics textbook; example was accurate.
Retrieval: Wrote definition and new example from memory.
Reconstruction: Created a decision matrix showing opportunity costs.
Transfer: Applied to a personal investment decision.
Reflection: I struggled with quantifying opportunity cost; need more practice.
Boundaries: I will use AI only for examples, not to calculate my own decisions.
Instructions: Use an AI system to generate an explanation of a concept you are learning (or use the supplied AI explanation below). Apply TRACE and CLAIM to verify it. Document your findings.
Supplied AI explanation: "The concept of 'elasticity' in economics measures how much demand or supply responds to changes in price. It is always negative because price and quantity move in opposite directions. For example, if the price of a good increases by 10%, demand will always decrease by exactly 10%."
Show analysis
TRACE review:
- Factual error: Elasticity is not always negative; it can be positive for inferior goods (Giffen goods). Also, the statement "always decreases by exactly 10%" is false; elasticity varies.
- Reasoning error: The AI conflates general principles with universal laws.
- Uncertainty: The explanation lacks nuance about different types of elasticity.
CLAIM verification:
- Check a microeconomics textbook: Elasticity is the percentage change in quantity demanded divided by percentage change in price; it can be negative, positive, or zero.
- Corrected explanation: "Price elasticity of demand measures the responsiveness of quantity demanded to a price change. It is usually negative for normal goods, but can be positive for Giffen goods. The magnitude varies."
Conclusion: The AI explanation contains significant errors. Do not learn from it without correction.
Instructions: Create a visual or written workflow that outlines the steps you will follow when using AI to support your learning. Include: learning goal, independent attempt, bounded AI assistance, verification, retrieval, reconstruction, transfer, reflection, documentation, human referral, and stop conditions. Make it specific to your typical learning contexts.
Show sample workflow outline
- Define learning goal.
- Attempt independently for at least 20 minutes.
- If stuck, ask AI for a hint or specific clarification (not the answer).
- Verify AI's contribution using TRACE and CLAIM.
- Close AI and retrieve the information from memory.
- Reconstruct the knowledge in my own words.
- Apply to a new problem (transfer).
- Reflect on what I learned and what is still unclear.
- Document AI use and my learning evidence.
- If still stuck, consult a human (instructor, tutor).
- Stop using AI if it starts generating answers instead of hints.
Instructions: Create a personal charter that specifies your boundaries for AI use in learning. Address: permitted practice uses, uses requiring verification, uses requiring disclosure, course-dependent uses, no-AI activities, protected information, human-support needs, and stop conditions.
Show sample charter (partial)
- Permitted: Brainstorming, generating practice questions, clarifying concepts, checking grammar.
- Verification required: All factual explanations, citations, data.
- Disclosure required: Any AI contribution to graded work (if allowed).
- No-AI: Personal reflections, examinations, unaided problem sets.
- Protected: Never share personal, health, financial, or confidential information.
- Human support: Consult instructor when course rules are unclear.
- Stop: If AI starts writing content for me, or if I cannot explain the output.
Instructions: Reflect on a recent learning activity where you used AI. Compare your recall, reconstruction, and transfer performance with and without AI (or with different levels of AI use). Assess whether the AI helped you learn or merely helped you complete the task.
Show sample evaluation
Activity: Learning to calculate net present value (NPV).
With AI: I used AI to generate a step-by-step solution for a practice problem. I then solved a similar problem on my own. I could solve it, but I wasn't sure why the steps worked.
Without AI: I attempted a problem from scratch, made errors, and corrected them. It took longer, but I understood the reasoning.
Evaluation: The AI helped me complete the task faster, but it did not deepen my understanding. I would have learned more by struggling through without AI, then using AI to check my work.
Conclusion: AI improved completion but not learning. I will use AI only for verification, not for step-by-step solutions.
Exercises
Complete these exercises to reinforce your understanding. Attempt each before reviewing the answer.
Task: You are learning about "Maslow's hierarchy of needs." Describe how you would apply the LEARN framework to this concept. Include the learning goal, independent attempt, AI assistance, verification, retrieval, reconstruction, transfer, and reflection.
Show solution design
Learning goal: Explain Maslow's hierarchy and apply it to a workplace scenario.
Independent attempt: Write the levels from memory and give an example for each.
AI assistance: Ask AI: "Give me a workplace example of how each level of Maslow's hierarchy can be addressed."
Verification: Check with a management textbook.
Retrieval: Write the hierarchy and examples without notes.
Reconstruction: Create a diagram from memory.
Transfer: Apply to a different setting (e.g., a classroom).
Reflection: Note which levels I struggled with and plan to study them more.
Task: Use the following AI-generated statement: "The average human brain uses 20% of the body's oxygen." Apply TRACE and CLAIM to verify this claim. Describe your findings.
Show answer
TRACE: The claim is a factual statement about oxygen consumption. It may be a simplification or may be inaccurate.
CLAIM: Check multiple reliable sources (e.g., medical textbooks, physiology articles). Most sources indicate that the brain uses about 20% of the body's oxygen, but this is an average and varies. The claim is generally accurate, but needs context.
Conclusion: Verified, but note that it is an average and may not apply to all individuals or conditions.
Task: List three learning activities that should be performed entirely without AI (no-AI activities). Explain why each should remain human-only.
Show sample answer
- Personal reflection essays: The purpose is to develop self-awareness and personal expression; AI cannot reflect your unique experience.
- Final examinations: Designed to assess your unaided knowledge and reasoning.
- Oral presentations: You need to develop communication skills; AI cannot deliver the presentation for you.
Task: Design a workflow for learning a new software tool (e.g., a project management app) using AI support. Include independent attempt, AI assistance, verification, retrieval, reconstruction, transfer, and reflection.
Show solution design
- Try to use the software without help (independent attempt).
- If stuck, ask AI: "How do I create a project timeline in this tool?"
- Verify the AI's instructions against the official documentation.
- Close AI and try to recreate the steps from memory (retrieval).
- Reconstruct the workflow as a checklist.
- Apply to a different project (transfer).
- Reflect on which features I still need to learn.
Task: You are in a course where the instructor allows AI for research assistance but not for writing. You are working on a research paper. Define what AI use is permitted, what requires verification, and what is prohibited.
Show sample boundary
Permitted: Generating search terms, suggesting sources (to be verified), checking grammar.
Verification required: All sources, facts, and citations provided by AI.
Prohibited: AI writing any part of the paper, generating the thesis, or producing analysis.
Disclosure: Must include a statement of AI use.
Task: Write a disclosure statement for a learning activity where you used AI to generate a list of potential topics and to get feedback on your outline. You did not use AI for the writing.
Show sample disclosure
Disclosure: "I used AI to generate a list of potential topics for this assignment and to provide feedback on my outline. All analysis, writing, and argumentation are my own. I verified all AI-suggested sources."
Task: You used AI to generate a summary of a chapter. You then wrote a summary from memory. Compare the two summaries. Did the AI help you learn, or did it merely save time?
Show sample evaluation
AI summary: Comprehensive, well-structured.
My summary: Missed some key points; showed gaps in understanding.
Evaluation: The AI summary helped me see what I missed, but I did not learn those points until I reconstructed them. The AI did not improve learning by itself; the reconstruction did. I will use AI to check my recall, not to replace my reading.
Task: For a learning activity of your choice, define at least three conditions that would trigger you to seek human assistance, and three conditions that would trigger you to stop using AI.
Show sample
Human referral conditions: 1) If AI gives contradictory answers. 2) If I am confused about course policy. 3) If I need clarification that AI cannot provide.
Stop conditions: 1) If AI starts generating assessed content. 2) If I cannot explain the AI's output. 3) If AI requests sensitive information.
Homework Problems
Task: Match each LEARN component (L, E, A, R, N) to its description.
- 1) Independent effort before AI.
- 2) Applying knowledge to new contexts.
- 3) Defining what you need to develop.
- 4) Using AI in a bounded way.
- 5) Testing recall, reconstructing, and verifying.
Show answers
1) E – independent effort.
2) N – next steps/reflection (including transfer).
3) L – learning goal.
4) A – appropriate AI assistance.
5) R – retrieval, reconstruction, verification.
Task: A student uses AI to generate a practice problem, solves it independently, then uses AI to check the solution. Analyse this use against LEARN. Which components are present? Which are missing?
Show answer
Present: Learning goal (solving the problem), independent effort (solving independently), appropriate AI assistance (generating the problem and checking), retrieval (solving without AI).
Missing: Verification (checking the AI's solution against a source), reconstruction (the student did not reconstruct the method), transfer (no application to a new problem), reflection (no evaluation).
Task: You are learning about "Pareto efficiency" in economics. Write a brief plan for how you would use TRACE and CLAIM to verify an AI-generated explanation of the concept.
Show solution design
Plan:
- Ask AI: "Explain Pareto efficiency."
- Use TRACE: check for errors or oversimplifications.
- Use CLAIM: compare with an economics textbook.
- Note any discrepancies and correct.
- Learn from the corrected explanation.
Task: Identify three learning tasks that should never be done with AI, even for practice, and explain why.
Show sample
- Developing a personal ethical stance: AI cannot reflect your values.
- Practicing a physical skill: You must physically perform it.
- Creating an original work of art: The act of creation is the learning.
Task: Design a workflow for learning to write a business case analysis. Include each step of the LEARN framework.
Show solution design
- L – Goal: write a case analysis that identifies problems and recommends solutions.
- E – Read the case and write initial notes independently.
- A – Ask AI for frameworks or examples (e.g., SWOT analysis).
- R – Verify the frameworks, then write analysis from memory.
- N – Apply to a different case; reflect on the process.
Task: Your instructor allows AI for brainstorming but not for writing the final paper. You use AI to generate a thesis statement. Is this permitted? Justify.
Show answer
Answer: It depends on interpretation. Brainstorming is typically idea generation; a thesis statement is a core argument. Many instructors would consider a thesis statement part of the writing process, so it may be prohibited. You should ask your instructor for clarification.
Task: Write a disclosure statement for a project where you used AI to generate a list of potential data sources and to check grammar, but all analysis and writing were your own.
Show sample
Disclosure: "I used AI to identify potential data sources for this project and to proofread my final draft for grammar. All analysis, interpretation, and writing are my own. I verified all sources suggested by AI."
Task: You used AI to summarise a textbook chapter. Your exam scores on that chapter were lower than on chapters you read without AI. What does this suggest about the effectiveness of AI for learning in this case?
Show answer
Answer: It suggests that using AI to summarise may have replaced your active reading and comprehension, leading to shallower learning. You may have recognised the content but not encoded it deeply. A more effective approach would be to read actively, then use AI to check your understanding.
Task: You are preparing for a professional certification exam that tests practical skills. Should you use AI for practice questions? Why or why not?
Show answer
Answer: Using AI to generate practice questions can be helpful, but you should not use AI to answer them. The practice must be your own. The no-AI boundary applies to the performance of the skill, not to the creation of practice materials.
Task: You are using AI to help you understand a difficult concept. The AI gives a clear explanation, but it conflicts with something your instructor said in class. What should you do?
Show answer
Answer: Seek human help immediately. Consult your instructor or TA to clarify the discrepancy. Do not rely on AI when it conflicts with authoritative human instruction.
Task: You are using AI for feedback on your draft. The AI gives feedback that is vague and overly positive. You also notice that you are copying the AI's suggestions without thinking. Should you stop using AI? Why or why not?
Show answer
Answer: Yes, you should stop. The feedback is not useful, and you are passively copying, which is not learning. You should seek human feedback or use AI only if you can critically evaluate its suggestions.
Task: Apply the LEARN framework to a professional development scenario, such as learning to use a new project management tool at work.
Show solution design
L – Goal: use the tool to track project milestones.
E – Try to set up a project without help.
A – Ask AI for a step-by-step guide (verify with official docs).
R – Close guide and set up a test project from memory.
N – Apply to a real project; reflect on what you learned and what you still need.
Mini-Project: Complete LEARN-Based AI-Supported Learning Plan
Integrate all Unit 12 learning into a comprehensive plan for a bounded learning activity. Estimated effort: 180–240 minutes (this is a substantial project).
Task
Select one bounded learning activity from your own studies or from the provided list. Develop a complete LEARN-based plan that includes the following 40 sections:
- Course or discipline
- Learning activity
- Learning goal
- Knowledge or skill to be developed
- Evidence of learning
- Prior knowledge
- Initial confidence
- Independent attempt
- Proposed AI role
- Permitted AI assistance
- Prohibited or unsuitable AI assistance
- PCAFE-V prompt (if applicable)
- Source and context boundaries
- AI-generated explanation or feedback
- TRACE review
- CLAIM verification
- Corrected learning resource
- Productive-struggle plan
- Graduated hints
- Scaffolding strategy
- Scaffold-fading plan
- Retrieval-practice plan
- Self-explanation task
- Independent reconstruction
- Near-transfer activity
- Farther-transfer activity
- Cognitive-offloading analysis
- Confidence-calibration record
- Accessibility considerations
- Privacy boundary
- Academic-integrity boundary
- AI-use disclosure
- Human-referral condition
- Stop conditions
- Learning-effectiveness measures
- Complete LEARN analysis
- Residual limitations
- Final recommendation
- Assignment 1 reflection connection
- Preparation for Unit 13
Provide a structured submission (e.g., a document or detailed outline).
Show staged hints
Hint 1: Start with a clear learning goal. Be specific about what you will be able to do after the activity. Use verbs like "explain," "apply," "evaluate," "create."
Hint 2: Design the AI assistance carefully. Distinguish between what is permitted and what is prohibited. Use the assistance-vs-substitution criteria from Lesson 12.1.
Hint 3: Plan for verification and transfer. Include specific verification steps (TRACE and CLAIM) and design both near and farther transfer tasks.
Show solution architecture and partial example (for learning "cognitive dissonance")
Partial example (first 11 sections):
- Discipline: Psychology
- Learning activity: Understanding cognitive dissonance theory.
- Learning goal: Define cognitive dissonance, explain its effects on behaviour, and identify it in real-world situations.
- Skill: Conceptual understanding and application.
- Evidence: I can explain the theory and give three original examples.
- Prior knowledge: Basic understanding of attitudes.
- Initial confidence: 60%.
- Independent attempt: Write a definition from memory and find one example.
- Proposed AI role: Assistant – to clarify the theory and generate alternative examples for verification.
- Permitted: Asking for clarification, generating practice examples, feedback on my definition.
- Prohibited: AI writing my definition or generating my examples for me.
This is a partial example. Your complete plan should include all 40 sections.
Final recommendations (allowed): Proceed as designed, Proceed with verification, Proceed with reduced AI assistance, Use hints rather than answers, Add human instruction or feedback, Redesign the learning activity, Reserve AI for practice only, Use a non-AI learning strategy, Pause AI use, Do not use AI for this activity.
- Did I clearly define the learning context and goal?
- Did I specify independent attempt and permitted/prohibited AI assistance?
- Did I include TRACE and CLAIM verification?
- Did I plan for retrieval, reconstruction, and transfer?
- Did I address cognitive offloading, confidence calibration, and accessibility?
- Did I define boundaries, disclosure, human referral, and stop conditions?
- Did I evaluate learning effectiveness and provide a final recommendation?
- Did I connect to Assignment 1 and Unit 13?
Self-Check Quiz
Test your understanding with these questions. Attempt each from memory before revealing the answer.
1. What is the primary purpose of the LEARN framework?
Show answer
Answer: LEARN provides a structured process for responsible AI-supported learning that preserves learner effort and responsibility.
2. Which LEARN component involves applying knowledge to a new context?
Show answer
Answer: D) Next steps/reflection includes transfer and reflection.
3. What is the first step in verifying an AI-generated explanation using TRACE and CLAIM?
Show answer
Answer: B) Identify the claims. You must know what the AI is saying before you can verify it.
4. Which of the following is a learner-only activity?
Show answer
Answer: B) Writing a personal reflection. It must come from your own experience and thinking.
5. What should you document when using AI for learning?
Show answer
Answer: B) Document the prompt, output, verification, and your reconstruction. This supports learning and transparency.
6. When should you seek human assistance instead of AI?
Show answer
Answer: B) When the AI's response conflicts with course material. Human expertise is needed to resolve conflicts.
7. Which of the following is a stop condition for AI use?
Show answer
Answer: B) The learner cannot explain the result. If you cannot explain it, you have not learned it.
8. What does it mean to evaluate whether AI improved learning rather than merely accelerated completion?
Show answer
Answer: B) Compare recall, reconstruction, and transfer performance. Learning is about durable knowledge, not speed.
9. What is the purpose of a Learning-Boundary Charter?
Show answer
Answer: B) To define permitted, verification-required, and prohibited uses, along with other boundaries.
Assignment Connection
The verification skills from Assignment 5 are fully integrated into this lesson. You must apply TRACE and CLAIM to all AI-generated learning material.
This lesson provides comprehensive evidence for Assignment 1, including: your LEARN walkthrough, your verification analyses (TRACE and CLAIM), your personal workflow and boundary charter, your learning-effectiveness evaluation, and your complete LEARN-based plan (mini-project).
Consider adding these to your journal, especially your reflection on how your responsible-use boundaries have evolved.
Assignment 1 Journal Connection
Consider the following prompts for your Assignment 1 journal entry. Reflect honestly on your own practices and experiences.
- Task completion vs. learning: Reflect on a time when you completed a task with AI but did not learn the underlying material. How did you feel? What did you learn from that experience?
- Assistance vs. substitution: How has your understanding of the distinction between AI-supported and AI-substituted learning changed throughout Unit 12?
- LEARN: How has the LEARN framework (or its operational components) changed your approach to using AI for learning?
- Boundaries: What new boundaries have you set for your AI use? What no-AI activities have you identified?
- Learning effectiveness: How do you now evaluate whether AI has improved your learning? What metrics do you use?
Show reflection-planning checklist
- Did you reflect on a specific instance of task completion without learning?
- Did you analyse how your understanding of assistance vs. substitution has evolved?
- Did you describe how LEARN has changed your practice?
- Did you identify new boundaries and no-AI activities?
- Did you evaluate whether AI has improved your learning, not just completion?
Key Takeaways
- LEARN provides a structured framework for responsible AI-supported learning: Learning goal, Independent effort, Appropriate AI assistance, Retrieval/reconstruction/ verification, and Next steps/reflection.
- TRACE and CLAIM are essential for verifying AI-generated content before you learn from it.
- A personal learning workflow ensures you maintain control and responsibility when using AI.
- Learning boundaries (charters) help you define what is permitted, what requires verification, and what is prohibited.
- Documentation and disclosure are not just compliance measures; they support learning and reflection.
- Learning-effectiveness evaluation focuses on whether AI helped you develop durable knowledge, not just complete tasks.
- Human referral and stop conditions are critical safety nets to prevent over-reliance on AI.
- Some activities must remain entirely human-led (no-AI) to preserve authentic learning.
- The skills developed in Unit 12 directly prepare you for Unit 13: Research and Scholarly Practice.
- Responsible AI use is a deliberate, reflective, and human-governed practice.
Key Terms
- LEARN framework
- A structured process for responsible AI-supported learning: Learning goal, Independent effort, Appropriate AI assistance, Retrieval/reconstruction/verification, Next steps/reflection.
- Learning boundary
- A rule that defines what AI use is permitted, what requires verification, what requires disclosure, and what is prohibited.
- Learning-Boundary Charter
- A personal document that codifies your boundaries for AI use in learning.
- Documentation
- The practice of recording AI prompts, outputs, verification, corrections, and learning evidence.
- Disclosure
- A formal statement acknowledging AI use in a learning activity or submission.
- Learning-effectiveness evaluation
- The process of assessing whether AI use improved learning (recall, reconstruction, transfer) or merely accelerated task completion.
- Human referral
- Seeking help from a human instructor, tutor, or other qualified person when AI is insufficient.
- Stop condition
- A signal that it is time to stop or limit AI use, such as when it begins performing assessed work or you cannot explain the output.
- No-AI activity
- A learning task that must be performed entirely without AI to preserve authentic skill development.
- Personal AI learning workflow
- A structured sequence of steps you follow when using AI to support your learning.
Looking Ahead
This concludes Unit 12. You have now developed a comprehensive set of skills for using AI responsibly to support your learning. In Unit 13: Research and Scholarly Practice, you will build on these skills to apply them in research contexts. You will learn to develop research questions, conduct literature searches, synthesise scholarly sources, adhere to research ethics, and use the RIGOUR framework — all with responsible AI use. The habits of verification, documentation, and boundary-setting you have cultivated here will be directly transferable.
Sources and Further Reading
- Primary source: COMP 3270 Textbook, Chapter 12: Generative AI for Learning (sections on LEARN, workflows, and boundaries). [PAGE NUMBER TO BE CONFIRMED]
- Supporting: Chapter 7: Hallucination, Error, and Uncertainty (TRACE). [PAGE NUMBER TO BE CONFIRMED]
- Supporting: Chapter 8: Verification, Evidence, and Epistemic Responsibility (CLAIM). [PAGE NUMBER TO BE CONFIRMED]
- Supporting: Chapter 13: Writing, Communication, and Creative Work (selected sections). [PAGE NUMBER TO BE CONFIRMED]
- Recommended: Appendix A: Generative AI Glossary; Appendix C: Verification Toolkit; Appendix D: Responsible-Use Toolkit; Appendix H: Sample AI Use Statements. [PAGE NUMBERS TO BE CONFIRMED]
[AUTHOR OR SUBJECT-MATTER EXPERT REVIEW REQUIRED]