Adult Education: AI Ethics & Academic Integrity Foundations

After completing this lecture, learners will be able to:

• Explain why AI text detectors can produce unreliable results, including particular concerns for multilingual writers.

• Evaluate detector scores critically rather than treating them as proof of AI authorship.

• Explain, in practical terms, how features such as predictable language and perplexity can affect detection.

After completing this lecture, learners will be able to:

• Explain why AI authorship detection alone is insufficient for making academic integrity decisions.

• Shift their evaluation from asking whether AI “touched” an assignment to examining evidence of whose thinking shaped the work.

After completing this lecture, learners will be able to:

• Introduce AI expectations to students in a way that emphasizes responsibility, judgment, and transparency rather than fear.

• Design a Day 1 AI conversation that helps students recognize both the capabilities and limitations of generative AI.

• Use the question “How do we make student thinking visible?” to guide later assessment decisions.

After completing this lecture, learners will be able to:

• Explain how AI detection and AI-evasion strategies can create an ongoing technological arms race.

• Identify the educational costs of a detection-first approach, including the Detection Tax on instructor time and attention.

• Recognize how surveillance can shift students from improving their writing toward trying to make their work appear human.

After completing this lecture, learners will be able to:

• Apply the four-step framework of Understand, Clarify, Make Visible, and Redesign to an existing course or assignment.

• Replace detection-first practices with designed accountability that makes student thinking and decision-making more visible.

• Identify one concrete change they can make to strengthen an assessment in an AI-rich environment.

What the tool is, why it makes things up, and what that means for your students’ thinking.
Download: the How-LLMs-Work handout, with the ten-minute class activity.
Try this before the next lecture. (2 minutes)
Which of the four sentences would most surprise your students, and why? Write one sentence you could say to them in your own classroom. No submission needed; keep it for the Struggle Audit at the end of the module.

A method your students can run on any AI answer in about a minute.
Download: the Find-the-Lie Activity Pack with the Critique Rubric.
Try this before the next lecture. Activity A: Build your Find-the-Lie task.
Open the Find-the-Lie Activity Pack. Choose one variant (Citation Hunt, Red-Pen Annotation, or Grade-the-Machine), name the course it is for, and write the two lines that adapt it: what your students will verify, and what counts as a fabricated, wrong-details, or distorted claim in your discipline. Optional: post in the Q&A which variant you chose and what you swapped in for the journal references.

What the research on AI and thinking shows, and what it does not.
Download: the Evidence One-Pager (four questions on top; one row per claim).
Try this before the next lecture. Activity B: Rewrite three headlines, then one self-explanation.
Rewrite each headline so it says only what the described evidence supports. (1) “Students who use AI remember less”: a small preprint, one essay task. (2) “AI tutoring doubles learning”: a randomized trial in one physics course, two lessons, expert-written prompts. (3) “Over-reliance on AI harms learning”: a systematic review of mostly non-causal studies. Then, in one sentence: why does the design of the help matter more than whether AI was used at all? Use the controlled study from the lecture.

The Developmental Lens and the Struggle Audit: which difficulty to protect, and where AI belongs.

Download: the Serve-Share-Substitute case sort (with key) and the Struggle-Audit worksheet (filled model and blank).

Try this before the next lecture. Activity C: Serve, share, or substitute? (case sort).

Download the case sort. Sort the ten AI uses before you look at the key, then check. Note the ones you got wrong: those are the near-misses. For one of them, write the constraint that would move it to “serve.”

Activity D: Run the Struggle Audit on one lesson you teach.

Use the Struggle-Audit worksheet on one real lesson: learning target, cognitive work, every difficulty, protect/share/remove, the boundary written for students, and the evidence you would ask for. Run the access and workload checks. Keep the page; you will turn one PROTECT row into a rule in Module 4 and redesign the task around it in Module 6.

Choose a level, explain why it fits the learning, and change the task so the level is more than a label.

Download: the AIAS Statement Templates (one model statement per level, and three blanks for your own).

Then complete Practice Assignment 1 below.

Using the templates, you will label three of your own assessments; the assignment item under this lecture has the full instructions.

Two pocket tools for the 11 p.m. decision: a two-question map and a five-question flowchart.
Download: the Target-Work Decision Map and the “May I Use AI?” Flowchart.
Try this before the next lecture. Build and narrate one case.
Take one real AI question your students have asked. Run it through the five flowchart questions out loud (record yourself or write it out). Then write the allowed alternative you would offer if the answer is no. Optional Q&A post: the case and your redirect.

Where AI entered the work, what it contributed, what the student checked, and which judgments still belong to the student.

Download: the 6Cs Protocol Pack with the AI Use Declaration Builder and the citation-versus-disclosure guide.

Try this before the next lecture. Write and stress-test an AI Use Declaration.

Using the Declaration Builder, write a declaration for one way you have used AI in your own teaching work. Stress-test it: does it identify the stage? show what you checked or changed? make responsibility clear? Model it before you require it.

A clear frame, real student voice inside it, and an agreement stress-tested against the situations students will face.

Download: the Agreement Facilitation Kit (five parts and a change log) and the Syllabus Language Bank.

Try this before the next lecture. Build your facilitation plan (this becomes part of your capstone).

Using the Agreement Facilitation Kit: write your fixed frame and negotiable space, choose the six scenario-sort uses for your course, adapt the three stress tests, and set a review date. Do not write the agreement alone; the plan is what you bring to class. Keep the change-log template ready for after the session.

After completing this lecture, learners will be able to:

• Explain why a final written product alone provides limited evidence of how student thinking developed.

• Identify appropriate forms of process evidence, such as drafts, version histories, notes, and revision decisions.

• Begin shifting assessment from final-product inspection toward evidence of learning development.

After completing this lecture, learners will be able to:

• Distinguish between AI completing a revision and AI providing feedback that the student evaluates.

• Apply the Manual Draft → AI Critique → Rebuttal Log → Revised Draft model to an AI-supported writing activity.

• Use a Rebuttal Log to make students' acceptance, rejection, and modification of AI feedback visible.

After completing this lecture, learners will be able to:

• Build a manageable writing portfolio using baseline work, developmental drafts, revision evidence, and final work.

• Treat discrepancies between earlier and later work as questions for investigation rather than automatic evidence of misconduct.

• Select enough process evidence to understand student development without creating a new Documentation Tax.

After completing this lecture, learners will be able to:

• Conduct a brief oral check using Process, Rationale, and Reflection questions.

• Use oral conversations to assess student understanding without turning them into interrogations or AI detectors.

• Interpret hesitation and difficulty cautiously, particularly when working with multilingual or otherwise diverse learners.

Name the claim behind the grade, separate secured evidence from open learning, and give AI a defined role inside the task.
Download: the Method Card and the Two-Lane Sorting worksheet.
Try this before the next lecture. Build your two-lane map.
Using the Method Card, write the learning claim behind one major assessment in one sentence (the capability, not the product). Then complete the Two-Lane Sorting worksheet: list every piece of evidence the assessment produces, place each in a lane with its AIAS level, and answer the three questions. If one take-home product is carrying the whole claim, name one repair: add, split, or rebalance.

Put one real assignment under pressure, find the target work, and choose the smallest repair that makes the evidence more trustworthy.

Download: the Assignment Audit Worksheet (seven steps).

Then complete Practice Assignment 2 below.

You will run the Would This Pass? test on one of your own assignments and post the before and after in the Q&A; the assignment item under this lecture has the full instructions.

Map one assessment onto pre-task, during-task and post-task; set an AI level for each; make every phase depend on the one before.
Download: the Three-Phase Planner and the Peer Evaluation Protocol.
Try this before the next lecture. Plan one three-phase task.
Using the Three-Phase Planner, map one assessment onto pre-task, during-task and post-task. For each phase write the learner action, the AIAS level, the evidence, and the hand-off to the next phase. Run the target check and the connection check. Then test the Peer Evaluation Protocol on a sample: if a reviewer could score it without pointing to evidence, repair the criteria.

Three design essentials, the fairness conditions, and an honest workload estimate for a ten-minute oral in your own course.

Download: the Oral Exam Kit (question pool, follow-up bank, rubric, fairness checklist, workload table, peer mock protocol).

Try this before the next lecture. Capstone component: build a ten-minute oral.

Using the Oral Exam Kit: write the learning claim; choose defense, role-play or structured debate; write three core questions from the evidence moves with at least two follow-ups each; define what students may bring and the AIAS level for preparation and for the performance; adapt the criterion rubric; complete the fairness checklist and the workload table. Run the mock oral with a colleague or a small student group, and record what you need to change. Optional rehearsal: the Udemy Role Play “Hearing the student think.”

After completing this lecture, learners will be able to:

• Evaluate an AI-supported activity by asking whether meaningful cognitive work remains for the learner.

• Match AI support to a specific learning need rather than teaching prompting as an isolated technical skill.

• Distinguish between AI that supports learning and AI that substitutes for the thinking students need to practice.

After completing this lecture, learners will be able to:

• Design instructional AI prompts using Role, Task, and Constraint.

• Select cognitive, compensatory, or metacognitive prompting strategies according to the learning objective.

• Add appropriate constraints that preserve meaningful work for the learner.

After completing this lecture, learners will be able to:

• Design reverse-prompting activities in which AI questions the learner rather than immediately generating an answer.

• Create a Tutor Mode prompt that encourages students to generate, explain, and develop their own ideas.

• Determine when reverse prompting is appropriate and when another form of AI support would better serve the learning objective.

After completing this lecture, learners will be able to:

• Guide students through the Extract → Organize → Draft sequence after an AI-supported interview.

• Design a deliberate transition from AI-supported dialogue to independent student writing.

• Distinguish diagnostic AI feedback provided after a student attempt from AI generation that replaces the attempt.

After completing this lecture, learners will be able to:

• Identify how AI language polishing can improve conventional accuracy while potentially altering meaning, cultural expression, or student voice.

• Guide students in comparing AI revisions by asking what was gained, what was lost, and who decided.

• Reframe AI-generated language as an option for evaluation rather than a verdict on what constitutes “good” English.

After completing this lecture, learners will be able to:

• Design AI scaffolds that make language tasks more accessible without removing the learner's meaningful contribution.

• Match AI support to learner needs using principles from the Zone of Proximal Development, comprehensible input, interaction, and pushed output.

• Create a competence-based fade plan that gradually increases learner responsibility as capability develops.

After completing this lecture, learners will be able to:

• Help students identify rhetorical moves and their functions within academic texts.

• Distinguish AI assistance with recognizing genre structure from outsourcing substantive intellectual decisions such as the thesis, research gap, or contribution.

• Design AI-supported genre-analysis activities that preserve human authorship and disciplinary judgment.

After completing this lecture, learners will be able to:

• Recognize patterns that may indicate unhealthy reliance on AI, such as repeated approval seeking or automatically replacing one's own language.

• Help students develop AI literacy that includes evaluating outputs, preserving their own judgment, and recognizing when AI support should stop.

• Design learning experiences that protect opportunities for human interaction, self-trust, and independent decision-making alongside AI support.

After completing this lecture, learners will be able to:

• Break an extended assignment into distinct stages where different forms of AI assistance may require different boundaries.

• Create a Working Agreement specifying permitted AI support, work reserved for the student, and minimal process evidence for each stage.

• Adjust the level of responsibility for establishing AI-use terms as learners develop greater academic independence.

After completing this lecture, learners will be able to:

• Distinguish between locating a source and verifying that it genuinely supports the claim for which it is being used.

• Apply a systematic source-verification process to AI-assisted research.

• Design expectations that keep source selection, verification, and evidentiary judgment visible during extended academic work.

After completing this lecture, learners will be able to:

• Distinguish between AI identifying a pattern and a researcher judging the significance of that pattern.

• Use an Interpretation Decision Log to document important analytical decisions and the reasoning behind them.

• Use AI to pressure-test an existing interpretation without allowing the system to supply the interpretation itself.

After completing this lecture, learners will be able to:

• Distinguish among materials, feedback, and assessment when deciding how generative AI can appropriately be used in their own teaching practice.

• Apply the principle of human judgment before AI assistance when providing feedback, separating decisions about a student’s work from support with wording or expression.

• Identify instructor responsibilities when using AI for consequential assessment, including maintaining professional accountability for grading decisions.

• Evaluate whether their use of student work with AI complies with institutional requirements for privacy, data protection, and approved systems.

• Create a brief, transparent disclosure explaining to students how generative AI is used in their course.