SW4 · TEACHER GUIDE
Facilitator’s Guide: AI Ethics Council
A ten-day evidence trail: investigate the system, hear the people, test the tradeoffs, and write rules someone can actually check.
Essential question
How should people decide what AI is allowed to do?
What students do
Students work as an ethics council rather than completing ten disconnected AI activities. The same casebook carries evidence from one documented decision through stakeholder analysis, a public hearing, a student-written code, and a final revision.
Teacher materials
Download the 56-slide teacher deck · Make a Google copy
AI Ethics Council Casebook
Use the same casebook for all ten days. Earlier evidence stays available during the hearing and code-writing days.
Download PDFDownload editable Word copy · Make a Google copy
Before Day 1
- Choose or vet a manageable case/source set for Day 1. Students need one system, one decision, and a documented effect, not an open-ended search for “AI ethics.”
- Build a small Day 4 verification set with real, altered/generated, misleading, and genuinely uncertain items. Preserve the original source for each item.
- Choose the common Day 5 decision scenario.
- Assign Day 7 roles: student, teacher, parent, technology company, and district or government decision-maker.
- Prepare the debate timer and response order for Day 8.
- Choose paper posters or shared documents for the Day 9 code and sticky notes or a digital response method for Day 10.
Daily flow and leave-up screens
| Day | Teach | Casebook | Deck route |
|---|---|---|---|
| 1 | Case intake and source quality | Pages 1–2 | Slides 1–11; leave 10 up |
| 2 | Recommendation targets and stakeholders | Page 3 | Slides 12–17; leave 17 up |
| 3 | Privacy, consent, errors, and safeguards | Page 4 | Slides 18–23; leave 23 up |
| 4 | Lateral reading and verification | Page 5 | Slides 24–28; leave 28 up |
| 5 | Stakeholder hearing and power map | Page 6 | Slides 29–33; leave 33 up |
| 6 | Jobs as tasks; human accountability | Page 7 | Slides 34–38; leave 38 up |
| 7 | Debate role, evidence, rebuttal, condition | Page 8 | Slides 39–43; leave 43 up |
| 8 | Public hearing and personal verdict | Page 9 | Slides 44–47; leave 46 up during debate |
| 9 | Five-part, testable AI code | Page 10 | Slides 48–51; leave 51 up |
| 10 | Gallery walk, revision, final synthesis | Pages 11–12 | Slides 52–56; leave 55 up |
Instructional boundaries
- Discuss systems and documented cases, not classmates or families. Students may pass on personal disclosure and still analyze the supplied scenario.
- Do not require faces, feeds, locations, messages, or other personal data.
- A debate role is not a statement of the student’s personal belief.
- Do not assess speaking confidence as evidence quality. Reading from prepared notes, a recorded statement, or serving as evidence tracker can preserve the learning goal.
- “AI is good” and “AI is bad” are not conclusions. Require the decision, people, evidence, tradeoff, and check.
Grading and Canvas
- Days 1–7 and Day 10 are 5-point evidence checks. Day 8 and Day 9 use 100-point rubrics.
- Day 8 accepts upload, media recording, or text entry. Other days use upload or media recording.
- Review both 100-point rubrics before Day 7. Adjust the rating levels before publishing if they do not match your grading practice.
- The Xello Biases and Career Choices item is an optional/flex career connection. It is not required to complete the Council casebook.
English-first EB supports
Each projected work screen and Canvas day includes a word bank, complete sentence stem, labeled example or model, and visible success criteria. Provide translated support or oral clarification when needed, but do not stack two full lessons into the same student view.
Teacher background: Code.org: Our AI Code of Ethics · MIT AI + Ethics for Middle School · NIST face-recognition study summary · Civic Online Reasoning: lateral reading