Unit 5 — How Do I Use AI Ethically and Responsibly?

Unit 5 · Act · KC-5 Ethical Digital Citizenship

How Do I Use AI Ethically and Responsibly?

“Knowing how to use AI is not the same as knowing when to use it, when to stop, and who is affected when you do. Technology does not make decisions — people do.”

2 Sessions · 100–120 min

8 Activities · Capstone

Grades 7–12

KC-5

Learning Outcomes

LO-1 · Map the potential harms of AI misuse using a structured case-study framework

LO-2 · Identify the Five Faces of AI Risk and give a real-world example of each

LO-3 · Explain the difference between misinformation and disinformation, with an example of each

LO-4 · Analyse an AI harm scenario and identify every stakeholder and their power to prevent it

LO-5 · Audit a real AI use policy against five criteria and propose a specific improvement

LO-6 · Write a personal Ethical AI Pledge with specific, observable commitments — personal and civic

Why This Unit Is the Capstone

Unit 5 is not a new topic added at the end. It is the ethical dimension of every skill taught in Units 1–4, made explicit and actionable. You are no longer only a user of AI — you become a digital citizen.

Big IdeaEvery AI decision is an ethical decision. Responsible use means asking who is affected — before you act, not after.

The Curriculum Bridge — From Skills to Citizenship

Unit	The Skill You Built	The Ethical Question It Opens
1 — Discover	I understand what AI is and what it can do	Who decides how AI learns? Whose voices are left out?
2 — Learn	I use AI to learn more effectively	Am I developing my own thinking — or outsourcing it?
3 — Think	I evaluate and verify what AI tells me	What happens when others don't? Who gets hurt?
4 — Create	I create with AI and disclose my process	What are my responsibilities for content others will consume?
5 — Act	I use AI as an ethical digital citizen	All of the above — now as a responsibility to the world

Keep this question in mind throughout both sessions

If AI makes everything easier — does that mean everything AI enables is right?

My answer NOW

Share with a neighbour

Talk to the person next to you. Did they answer yes, no, or “it depends”? What does their answer depend on?

You will return to this

At the end of Session 2 you will answer this question again — after writing your Ethical AI Pledge. Compare the two answers.

A question with no clean answerThis unit does not have a single correct position. Ethical literacy is not about knowing the right answer — it is about developing the habit of asking the right questions.

The Capstone Bridge — From Skills to Citizenship

Hook

See how every skill from Units 1–4 opens an ethical question you must now answer.

10 min · Session 1 · Engage

Why This Activity?

The hook must do two things: remind students how far they have come across the curriculum, and surface the ethical questions that skills alone cannot answer. This is the transition from user to citizen. Aim for productive moral tension — not a lecture. Let students disagree; the discomfort is the entry point.

You are no longer only a user

Four units ago you had heard of AI. Today you can explain it, prompt it, verify it, and create with it. But every one of those skills opens a question that the skill itself cannot answer: not can I do this? — but should I, and who is affected if I do?

Discussion — 3 minutes

1
Look at the “Ethical Question” column on the Overview page. Which question feels most urgent to you right now?
There is no right choice here. What matters is being able to say why that question matters to you.
2
Have you ever been in a situation where someone else's AI use affected you — without your knowledge or consent?
You can keep it general. A photo, a message, a recommendation, a grade — AI is involved in more of your day than you notice.
3
Is there a difference between “AI made this happen” and “a human decided to do this with AI”?
This is the central claim of the whole unit: technology does not make decisions. People do — and people are responsible for the decisions technology enables.

My answer to the third question

Watch For…

• Students who say “it depends how you use it” — excellent: push them to name a case where use causes harm even without bad intent. • Students who dismiss AI ethics as irrelevant to them — ask: “Have you ever seen a photo online that turned out to be AI-generated? How did it make you feel?” • Students who immediately want to talk about AI taking jobs — note it as a real concern for Activity 3, and redirect to the bridge table for now.

AI Harm Mapping — Case Study Analysis

Explore

Analyse a realistic AI harm scenario: who is harmed, how, and what could have prevented it.

20 min · Session 1 · Group work

Why This Activity?

Students must encounter the ethical dimensions of AI through concrete, realistic scenarios — not abstract principles. Case-based learning activates moral reasoning in ways definitions cannot. The Harm Map makes the systemic nature of AI harm visible: harm is rarely limited to one person.

SetupGroups of 4–5. Each group takes one Case Study Card. Three colours of sticky notes: yellow = who is harmed, red = how they are harmed, blue = what could prevent it.

The Six Case Study Cards

Card	Scenario	Core Issue
A	A student uses AI to generate a fake voice message from a classmate and shares it as a joke. The classmate is humiliated and the “message” spreads to other schools.	Deepfake / Privacy / Digital harm
B	A news website uses AI to generate 50 articles a day. One contains fabricated statistics about a local health crisis. Parents share it widely. A hospital is flooded with calls.	Misinformation at scale / Accountability
C	A company's AI hiring system consistently ranks candidates from certain postcodes lower — areas that happen to be majority low-income. The company claims the AI is neutral.	Algorithmic bias / Discrimination
D	A platform's AI recommendation system learns that outrage keeps users engaged longer. It begins consistently surfacing content that makes users angry — across all topics.	Engagement-driven harm / Manipulation
E	A student submits an AI-generated essay as entirely their own. Their school has no AI policy. Their teacher has no way to know. The student is praised and given an award.	Academic integrity / Institutional gap
F	An AI image generator is trained on artworks scraped from living artists without consent. Artists find their distinctive styles replicated by tools available commercially to anyone.	Intellectual property / Consent / Displacement

Group Task — 12 minutes

1
Step 1 — Who is harmed?
List every person or group affected, directly or indirectly. Think beyond the obvious victim: bystanders, families, other users, the wider public.
2
Step 2 — How are they harmed?
For each group, name the specific harm: reputational, financial, psychological, educational, or loss of opportunity.
3
Step 3 — What could have prevented this?
Think in three levels: what the individual could have done, what the platform could have designed, and what the institution could have required.
4
Step 4 — Write your headline
“AI enabled ______ to happen, resulting in ______ for ______.” A headline forces you to name the harm precisely.

My group's headline

Gallery Walk — 5 minutesGroups rotate to read each other's Harm Maps. Each student adds one “+” sticky to the harm they consider most important. Return to seats; each group reports its headline and one prevention measure.

The harm I had not considered before the Gallery Walk

Watch For…

• Groups that focus only on the individual bad actor — expand: “Who else was harmed? Who had power to prevent this at a systemic level?” • Groups that answer “just ban AI” — probe: “Is that realistic? What specific rule, design choice, or behaviour would have helped?” • Groups who minimise the harm because it happened to young people — take it seriously: “Would this be less harmful if it happened to an adult? Why?”

Five Faces of AI Risk

Explain

Learn a framework for recognising AI harms — so you can spot them outside the classroom too.

10 min · Session 1 · Explain

Why This Activity?

After encountering specific cases, students need a conceptual framework to organise and recognise AI harms beyond the examples they have seen. The Five Faces model is memorable, transferable, and rooted in real policy discourse. Teach each Face with one real example — never as a definition alone.

Face	Name	Definition	Real-World Example
1	Misinformation & Disinformation	AI generates or amplifies false content — whether by accident or by design	AI-written news articles with fabricated quotes; viral health misinformation
2	Deepfakes & Identity Manipulation	AI creates convincing fake media of real people — image, video, voice	Political deepfakes; fake celebrity endorsements; cloned student voices
3	Algorithmic Bias & Discrimination	AI systems encode and amplify human bias in decisions that affect real lives	Hiring algorithms that discriminate; facial recognition errors; credit scoring
4	Privacy & Surveillance	AI enables collection, analysis and use of personal data at unprecedented scale	Behaviour tracking on social media; AI-powered cameras; data scraping
5	Displacement & Dependency	AI replaces human roles and can create reliance on systems we no longer understand	Creative displacement; job automation; decisions outsourced to AI

Misinformation vs. Disinformation — intent is the differenceMisinformation can spread without anyone meaning harm. Disinformation requires deliberate deception. The same false article can be both — depending on who shared it and why.

Check for Understanding

1
Which Face did your Case Study Card represent? Could it be more than one?
Most real cases involve two or three Faces at once. Naming all of them is the more accurate answer.
2
Which of the Five Faces do you meet most often in your own daily life?
Every recommended video, targeted advert, and filtered search result is an AI decision made about you.
3
Is there a Sixth Face this framework is missing?
Define it, give one example, and explain why it deserves its own category rather than fitting inside an existing Face.

My proposed Sixth Face

Watch For…

• Students who conflate misinformation and disinformation — clarify that intent is the distinction. • Students unaware they already use AI-powered systems — surface it gently. • Students who feel overwhelmed — normalise: “The point is not to fear AI. The point is to be a citizen who understands what it can do — including what it can do wrong.”

Who Gets Hurt? — Stakeholder Analysis

Elaborate

Move from identifying harm to analysing who bears it — and who had the power to stop it.

15 min · Session 1 · Elaborate

Why This Activity?

Moving from identifying harm to analysing who bears it — and who has the power to prevent it — is the shift from moral awareness to ethical agency. Stakeholder Analysis builds the mental habit of asking “who else?” before acting.

The most important ethical questionIt is not only “what went wrong?” but “who had the power to stop it — and why didn't they?”

Stakeholder Group	Relationship to the Harm	Power to Act
The Individual User	Often the immediate agent — makes the AI-enabled choice	High: can choose not to use AI in that way
The Platform / Developer	Designs the system — determines what it can and cannot do	Very high: can build safeguards, policies, limits
The Institution	School, company or government — sets the rules of AI use in its context	High: can create policy and enforce standards
The Affected Person	Bears the harm — may or may not have consented	Low in the moment; can advocate for rights afterwards
Society	Experiences the cumulative effect of many individual AI choices	Acts through collective decisions: law, norms, culture

The Two Hard Questions

1
Which stakeholder had the MOST power to prevent this harm — but did not use it? Why not?
Power and responsibility are not the same as blame. Someone can be blameless and still have had the power to prevent harm.
2
Is it fair to blame the individual user when the platform was designed to enable the harm?
There is no settled answer. Strong responses acknowledge both levels: individual choice inside a designed system.

My advocacy statement — one message to one stakeholder group

Watch For…

• Students who only blame “bad people” — expand to the systemic level. • Students who feel powerless (“we can't change anything”) — redirect: “The most powerful thing you can do today is make a deliberate choice about how you use these tools.” • Students who identify only the immediate victim — probe: what about those who shared the content? Other users of the platform?

Exit Ticket — Session 1 (D1)

5 minutes — individual and honestThis is not graded. It helps your teacher plan Session 2.

Name THREE of the Five Faces of AI Risk and give one concrete real-world example for each.

In your own words: what is the difference between misinformation and disinformation? Why does the distinction matter?

Complete and explain: “The stakeholder with the MOST responsibility to prevent AI harm is ______ because ______.”

One question I still have about AI ethics going into Session 2:

Teacher Use of This Exit Ticket

Identify students who answered Q3 with only “the user” — they need the systemic framing reinforced in Session 2.

Real Cases, Real Consequences

Warm-Up

Connect the systemic harms of Session 1 to three documented cases that actually happened.

8 min · Session 2 · Engage

Why This Activity?

Session 2 opens the ethical agency half of the unit. Students need to connect the systemic harms they studied to real, documented consequences before they can meaningfully commit to personal principles. These are not predictions about the future — they happened, and they affected real people.

Case	What Happened	The Consequence
Fabricated Legal Citations	A lawyer submitted AI-generated court citations that did not exist. Nobody checked them before submission.	The lawyer faced disciplinary proceedings. The case was damaged. The opposing party was harmed.
AI Voice Cloning Fraud	Scammers cloned a CEO's voice with AI and called a finance employee, who transferred funds to them.	Hundreds of thousands lost. The employee was not at fault — the technology made the deception convincing.
Exam Essay Accusation	A student was accused of using AI to write an exam essay. The detection tool was wrong — the student had written it.	The student faced a grade penalty and real distress. The tool had a known high false-positive rate.

Debrief — 3 minutes

1
In each case — was the harm caused by AI, or by a human decision?
In all three, a human chose not to verify, or chose to deceive, or chose to trust a tool beyond its reliability.
2
Which of the Five Faces does each case represent?
Case 1: misinformation. Case 2: deepfake and identity manipulation. Case 3: dependency on a system we do not fully understand.
3
Which case concerns you most — and why?
Most students choose the third, because it is the one that could happen to them. That instinct is worth naming out loud.

Bridge to Session 2“These are not predictions. They happened. Session 2 is about what YOU will do — as a user, a student, and a citizen — when you meet these moments.”

Watch For…

• Students who say “just don't use AI” — acknowledge the impulse, then probe: “Is that realistic? What principles would guide you when you do choose to use it?” • Students who dismiss the third case (“the student should have been more careful”) — redirect: “Is it fair to blame a student for a tool's error? Who should be accountable?”

AI Policy Audit — Reading the Rules

Explore

Read a real AI use policy critically, judge it against five criteria, and propose one improvement.

15 min · Session 2 · Explore

Why This Activity?

Ethical citizenship requires not just personal principles but policy literacy — the ability to read, evaluate and advocate around the rules that govern AI use in real institutions. This is a skill students will need in school, work and civic life.

What is an AI Use Policy?A document — formal or informal — that defines how AI may be used in a specific context: a school, a company, a platform, or a government. Reading these critically is part of AI literacy.

Audit Question	What to Look For
1. What does it PERMIT?	Which AI uses are explicitly allowed? Are any actively encouraged?
2. What does it PROHIBIT?	Which uses are explicitly banned? Are the prohibitions clear enough to follow?
3. Who is PROTECTED?	Does it protect users? Workers? Children? Everyone — or only the institution?
4. Who is ACCOUNTABLE?	If something goes wrong, who is responsible? Is that stated clearly, or left vague?
5. What is MISSING?	Which situations does the policy fail to address that it should?

My Verdict

Choose one policy to auditYour school's AI policy, a social media platform's AI content policy, or a national AI framework. If your school has no AI policy — that absence is itself your finding.

The policy I audited — and my rating (Strong / Adequate / Weak / Inadequate)

My proposed addition — one sentence, written exactly as it would appear in the document

Watch For…

• Students who conclude “there is no policy, so nothing applies” — this is the most important teachable moment in the activity: a decision NOT to have a policy is itself a policy choice with real consequences. • Students who audit only what a policy prohibits — push them to question 3 and 4: protection and accountability are where most policies are weakest.

My Personal Ethical Framework — The Five Lenses

Explain

Find the ethical lens that already guides you — and the one that challenges you most.

15 min · Session 2 · Explain + Elaborate

Why This Activity?

Students cannot make a genuine commitment from principles they have not personally examined. This activity forces self-knowledge before pledge-writing: which lens do I already use, and which one do I avoid? A pledge written without this step becomes a list of what the teacher wants to hear.

Lens	Core Question	Applied to AI Use
Harm Prevention	Does this cause harm to someone?	Before I use AI this way, could it hurt anyone — including someone not in the room?
Fairness	Is this fair to everyone involved?	Does this give me an unfair advantage? Does it treat all people equally?
Honesty & Transparency	Am I being honest about this?	Am I disclosing my AI use when it matters? Am I presenting AI output as my own?
Respect for Persons	Does this respect people's dignity and autonomy?	Would the affected person consent to this if they knew?
Responsibility & Impact	Am I taking responsibility for the outcome?	If AI helps me do this and something goes wrong, am I willing to be accountable?

The Five Lenses are not a checklistPassing three lenses does not make an action ethical. Each lens reveals a different dimension — and sometimes two lenses point in opposite directions. Naming that tension honestly is stronger than pretending it does not exist.

My default lens — the one that already matches how I think, with a real example from my own AI use

My challenge lens — the one where my instinct has sometimes led me away from the right choice

Watch For…

• Students who choose the lens that sounds most virtuous rather than most personally true — the whole activity fails if this goes unchallenged. Ask: “Give me an example from your own life where you actually used that lens.” • Students who claim all five lenses guide them equally — press for a ranking; refusing to prioritise is a way of avoiding the choice.

The Ethical AI Pledge — Capstone Commitment

Capstone

Write the final artifact of the whole curriculum: your own commitments, in your own words.

15 min · Session 2 · Elaborate

Why This Activity?

This is the capstone artifact of the entire curriculum. Its value is not in the elegance of the writing but in the specificity of the commitments. Model your own pledge first — including one commitment you find genuinely difficult. Students write honestly only when the teacher goes first.

This is a personal documentBe honest, not impressive. A genuine pledge that admits difficulty is worth far more than a perfect-sounding list of rules.

Section	Content	What Makes It Strong
1. My AI Story	2–3 sentences: how has your relationship with AI changed across this curriculum?	Specific — names a moment or realisation from a particular unit
2. What I Now Know	Three commitments: what I will ALWAYS do / what I will NEVER do / what I will ASK before acting	Concrete — names specific behaviours, not vague intentions
3. My Ethical Lens	Which of the Five Lenses will guide you most — and why?	Personal — reflects real self-knowledge, not the “right” answer
4. My Responsibility to Others	One commitment that goes beyond yourself — to your class, school, community or society	Civic — moves from personal ethics to citizen responsibility
5. My Pledge Signature	I commit to this framework as a starting point — and to revisiting it as AI and I both change.	Humble — acknowledges growth rather than claiming perfection

Write Your Pledge

1. My AI Story

2. What I will ALWAYS do · NEVER do · ASK myself

3. My primary ethical lens — and why it is the right one for me

4. My civic commitment — what I will do that protects or benefits others

5. The commitment I know will be HARDEST to keep — and the one I am MOST certain about

Watch For…

• Pledges that could belong to anyone (“I will use AI responsibly”) — send them back with one question: “In which specific situation, and what exactly will you do?” • Pledges with no civic section — the most common gap; every pledge needs one commitment that reaches beyond the student's own behaviour. • If a pledge reveals intent to cause harm, treat it as a pastoral matter, not an assessment one.

Exit Ticket — Session 2 (D2)

7 minutes — the final Exit Ticket of the curriculumAnswer all four questions individually and honestly.

Name ONE commitment from your Ethical AI Pledge you feel genuinely committed to — and one situation where it will be hard to keep.

Which of the Five Faces of AI Risk is most urgent for society to address right now? Why?

Name ONE thing missing from the AI policy you audited — and the harm that gap leaves unaddressed.

The final question of the curriculum: “The most important thing this curriculum taught me about AI is ______.”

From Session 1 to Session 2 — a shiftIn Session 1 you mapped AI harms. In Session 2 you committed to action. Before Session 1, when I thought about AI ethics I thought ______ . After Session 2, I now think ______ .

Concept Flashcards — Unit 5 Summary

How to useTap a card to flip it. Try to recall the definition before you flip — that pause is the real practice.

0 / 8 cards flipped

📰

Misinformation

Tap to flip

False or misleading information — not always deliberate, but still harmful when spread.

🎭

Disinformation

Tap to flip

Deliberately false information designed to deceive. AI makes it easier to create and spread at scale.

👤

Deepfake

Tap to flip

AI-generated media that makes a real person appear to say or do something they never did.

⚖️

Algorithmic Bias

Tap to flip

When an AI system produces unfair outcomes because of biased training data or design choices.

🔒

Digital Privacy

Tap to flip

Your right to control what information about you is collected, stored and used by AI systems.

🤖

Digital Displacement

Tap to flip

The replacement of human workers, voices, or creative output by AI systems.

📜

AI Policy Literacy

Tap to flip

The ability to read, evaluate and advocate around the rules that govern how AI may be used.

🛡️

Citizen Responsibility

Tap to flip

Your obligation as a member of society to consider how your AI use affects other people.

Why This Activity?

Flipping the card creates a half-second retrieval gap. That gap is where memory consolidates.

Watch For…

Students who flip every card in ten seconds without recalling anything. Set the rule aloud: think first, then flip. Use this panel as a two-minute revision race, not as reading material.

Unit Quiz — D3

8 questions — choose the best answerYou get immediate feedback on every option, right or wrong. Read the feedback even when you are correct — it explains why the other options fail.

Question 1

Which of the following best describes disinformation — as distinct from misinformation?

AFalse information that spreads accidentally because people misunderstand it

BInaccurate AI output caused by training data errors

CDeliberately false information created and spread with the intent to deceive

DInformation that is outdated rather than intentionally false

Question 2

A platform's AI learns that outrage keeps users engaged longer, and begins surfacing content that makes users angry. Which Face of AI Risk is this?

ADeepfakes & Identity Manipulation

BMisinformation & Disinformation

CAlgorithmic Bias & Discrimination

DDisplacement & Dependency

Question 3

In a Stakeholder Analysis, which group typically has the MOST power to prevent AI harm through design choices?

AThe Individual User

BThe Affected Person

CSociety

DThe Platform / Developer

Question 4

A student audits their school's AI policy and finds it says nothing about who is responsible if AI detection tools produce false accusations. Which Audit criterion has failed?

AWhat does it permit?

BWho is accountable?

CWhat does it prohibit?

DWho is protected?

Question 5

Which of the Five Ethical Lenses asks: “Would the affected person consent to this if they knew?”

AHarm Prevention

BFairness

CHonesty & Transparency

DRespect for Persons

Question 6

An AI hiring system consistently ranks candidates from low-income postcodes lower, even when qualifications are equal. This is:

AMisinformation & Disinformation

BDisplacement & Dependency

CAlgorithmic Bias & Discrimination

DPrivacy & Surveillance

Question 7

Which of these is a CIVIC commitment in an Ethical AI Pledge — as distinct from a personal one?

A“I will always write an AI Disclosure Statement before sharing AI-assisted work.”

B“I will never submit AI-generated content as entirely my own without disclosure.”

C“I will advocate for my school to develop a fair AI use policy that protects students.”

D“I will apply all Five Ethical Lenses before using AI in any significant situation.”

Question 8

Which statement about the ABSENCE of an AI use policy is most accurate?

AIt means AI use is unrestricted but also unprotected

BIt is neutral — it neither helps nor harms anyone

CIt protects individual users by giving them freedom to decide

DIt only matters in professional, not educational, settings

Watch For…

Question 2 has a defensible case for more than one answer. If a student argues for B or C with a clear reason, award the mark — the reasoning is the assessed skill, not the letter. Say so before they start; it models exactly the ethical uncertainty this unit teaches.

Ethical AI Pledge Rubric — D4 (Capstone)

Share this rubric BEFORE students write their PledgeClick the column that best describes your Pledge for each dimension. Do NOT grade whether a student reaches the “correct” ethical position — grade the quality of the ethical reasoning.

Pledge score

0/ 20 points

Dimension	4 · Exemplary	3 · Proficient	2 · Developing	1 · Beginning
1. Curriculum Synthesis	Connects to at least 2 prior units by name; cites a specific skill, insight or moment	References prior units, but the connection is surface-level or names only one unit	Mentions the curriculum in general terms without connecting to specific units	No connection to the curriculum at all
2. Specificity of Commitments	Every commitment names an observable behaviour in a named situation	Most commitments are specific; one is vague but the direction is clear	Partly specific, relying on vague intentions; no situation named	“I will use AI responsibly” — no behaviour, no situation
3. Ethical Lens Articulation	Names the lens, applies it to a concrete scenario, and admits where it creates tension	Names the lens and applies it to an example; no tension acknowledged	Names a lens without applying it, or picks the most virtuous-sounding one	Lens absent, or chosen as a formality with no reasoning
4. Civic Responsibility	At least one commitment names a specific impact on others: who benefits, how, and what the student will do	Civic commitment present but general; benefit named, action vague	Covers only the student's own AI use; no responsibility beyond avoiding wrongdoing	No civic dimension at all
5. Authenticity & Honesty	Names at least one genuine difficulty or past behaviour the student is committing to change	Feels honest, but no difficulty acknowledged; all commitments framed as easy	Performative — safe, correct-sounding statements with no evidence of real reflection	Copied from the model or so generic it could belong to anyone

Total	Level	Descriptor
18–20	Exemplary	Deep ethical reasoning, genuine personal commitment, specific civic responsibility, clear curriculum synthesis
14–17	Proficient	Solid reasoning with clear commitments; one or two dimensions lack depth
9–13	Developing	Engages with ethical ideas but commitments are vague, or the civic dimension is absent
4–8	Beginning	Generic, performative, or disconnected from the curriculum and the student's own AI use

A Note on Grading Ethics

The rubric rewards reasoning quality, not moral conformity.

Performance Task — Gate C (D6-C)

The final gate — no product is published before it passesGate B measured the product. This gate measures what happens when the product leaves the classroom and reaches people. Mandatory order: clearance first, then publication, then evidence, then defence.

Component	What It Measures	Time	Weight
Publication Clearance Form	Responsibility before action	10 min — before publishing	20
Creator's Defence	Authenticity under questioning	3 min per student	50
Impact Evidence	Reading the audience response	3–5 days + 10 min analysis	30

Publication Clearance — three checks

Area	What Must Be True Before Publishing
A · Accuracy	Every figure or scientific claim has a documented source; nothing taken from AI without independent verification; headlines do not overstate the evidence
B · Consent & Privacy	Everyone whose voice or image appears has explicitly agreed; no personal data without permission; parental consent if a minor appears; the requesting party approved the final version
C · AI Honesty	No generated images implying real people, places or events; a disclosure statement attached in a form suited to the channel; all sources and rights lawful and attributed

Stop signalA single “✗” halts publication until it is resolved. A student who finds a problem themselves and fixes it before publishing deserves a HIGHER mark than one who published without review. Publishing before the form is signed voids this component entirely — however good the product.

Creator's Defence — 3 minutes, unseen questions

How it runsOne minute of presentation, two minutes of questions. Panel: the teacher plus two students from another group. Three questions drawn at random from the bank — never given in advance. The student may consult their product and revision log; a prepared answer sheet or any AI tool is not allowed.

1
Why did you choose this particular form — the colour, the voice, the interface? What alternative did you reject, and why?
The answer must be about this audience, not about general good practice.
2
Show me one element of your work that would not exist if you had not made it.
This is the authenticity question. A student who cannot answer it has not made the work their own.
3
What was the last edit you made to an AI output — and why did you not leave it as it was?
Editing decisions reveal judgement. “It sounded wrong” is a start; ask what specifically was wrong.
4
If your audience changed — teachers instead of students — what would you change?
Tests whether the student designed for an audience or simply produced something.
5
What piece of information did you check and find inaccurate? What did you do with it?
The strongest answers describe a correction the audience never saw. That is Unit 3 becoming a habit.

A note on nervesHesitation is not weakness — some students do not speak fluently. The measure is the content: does the student know why they did what they did? A student who cannot answer any question has a work-ownership problem, not a speaking problem.

Impact Evidence — three real responses

Not likes. Not views.Likes and views reward excitement, not quality — a dangerous measure for teenagers. Collect three genuine responses from the target audience, record exactly what was said, and analyse what it means about your work. If nobody responds, that is also a result: analyse why. Was it the channel, the timing, or the message?

The one change I will make based on these responses

Full Curriculum Reflection — D5

The final document of the AI Literacy CurriculumComplete this after submitting your Ethical AI Pledge. Take your time — there is no right answer. Your teacher will keep this on file: it is the evidence of your complete learning journey.

Part 1 — What You Could Do Then, and Now

Unit	When You Started	What You Can Do Now
1 — Discover	You had heard of AI	You can explain what AI is and is not, and what it can and cannot do
2 — Learn	You used AI by instinct	You can design intentional prompts and use AI as a learning partner
3 — Think	You accepted or rejected outputs at random	You can verify, categorise and critically evaluate what AI tells you
4 — Create	You produced without a clear framework	You can author AI-assisted work and disclose your process with integrity
5 — Act	You used AI as a tool	You use AI as a citizen — aware of the impact on others

Part 2 — The Journey

1. Which unit changed how you think the most? What specifically shifted?

2. Which activity — across all five units — do you remember most vividly? Why did it stay with you?

3. Was there a moment where you felt uncomfortable? What caused it — and was it productive?

4. Describe one moment where you changed your mind about something — and what changed it.

5. Did anything here change how you use AI right now — today, outside of class?

Part 3 — Your Final Sentence

Complete this — the final sentence of the curriculum

If you could tell one thing to a student who has not yet taken this curriculum…

What would it be?

We return to our Essential Question…

If AI makes everything easier — does that mean everything AI enables is right?

Your answer at the start

Go back to the Essential Question panel and read what you wrote before Session 1. Has it changed?

Your answer now

The end of the curriculum — and the beginning of the question

You began this curriculum by asking: What is AI? You end it by asking: What kind of AI user do I want to be? The first question has answers. The second is something you will answer, every day, for the rest of your life. The work you did here is the beginning of that answer — not the end.

End of Unit 5 · End of the Curriculum

Teacher Guide · Unit 5 · Package A

Lesson Flow Notes

Timing and facilitation guidance across both sessions of the capstone unit.

Sensitive topics — read before you teachUnit 5 engages with real AI harms. Deepfakes, privacy violations and algorithmic discrimination may touch students' own experience. Create a safe space for honest discussion without requiring personal disclosure. If a student names a personal experience, thank them and move to the systemic level: “That is exactly why these rules and responsibilities matter.”

Session 1 · 0–10 min · Activity 1

Display the Capstone Bridge table. Give students two minutes of silent reading before any discussion — the table has to land before it is talked about. Open with: “You are no longer only a user of AI. Today you become a digital citizen.” Do not define ethics; let the questions in the right-hand column do the work.

Session 1 · 10–30 min · Activity 2

Groups of 4–5, one Case Study Card each. Distribute sticky notes in three colours before you explain the task — hands busy, attention ready. 12 minutes of group work, then a 5-minute Gallery Walk. Keep the walk moving: 60 seconds per map, then rotate.

Session 1 · 30–40 min · Activity 3

Teach each of the Five Faces with one real example — never as a definition alone. Two minutes per Face maximum. Then ask each group which Face their card represented; most will name two, which is the correct and more sophisticated answer.

Session 1 · 40–55 min · Activity 4

Introduce the Stakeholder Lens, then send groups back to their own case with one instruction: find every group, not just the victim. Close with the blame question as a whole-class debate — it has no settled answer and that is the point.

Session 1 · last 5 min · Exit Ticket D1

Individual and silent. Read the Q3 answers the same evening: students who wrote only “the user” need the systemic framing reinforced in Session 2. Two strong answers naming platforms or institutions are the best possible opening for Session 2 — read them aloud anonymously.

Session 2 · 0–8 min · Activity 5

Present the three real cases in one or two sentences each — resist elaborating. The third case (the wrongly accused student) is the one that lands hardest; leave the longest silence after it.

Session 2 · 8–23 min · Activity 6

Have the policy extracts printed in advance. If your school has no AI policy, use that: the absence is the strongest teaching material in the activity. Individually or in pairs; the five Audit Questions are answered in writing, not discussed.

Session 2 · 23–38 min · Activity 7

Introduce the Five Lenses in five minutes, then give ten to the two personal questions. State clearly that the lens they choose should be the one that is true for them, not the one that sounds best. Model this by naming your own challenge lens first.

Session 2 · 38–53 min · Activity 8

Share the D4 rubric before they write, not after. Model your own pledge — including one commitment you find genuinely difficult. Students write honestly only when the teacher goes first. This is quiet individual work; do not rush it.

Session 2 · last 7 min · Exit Ticket D2 + D5

D2 in class; D5 (the Full Curriculum Reflection) may go home if time is short. Keep D5 on file — it is the richest evidence of long-term learning synthesis in the entire curriculum.

Materials Checklist

Item	Quantity	Required For
Student Workbook (Package B)	1 per student	All activities
Case Study Cards	1 set per group (4–6 sets)	Activity 2
AI policy samples	2–3 printed or projected	Activity 6
Sticky notes, 3 colours	1 pad per group	Activity 2 Harm Map
Large paper or board space	1 per group	Harm Map display
Pledge rubric (D4)	1 per student before Session 2	Activities 7–8

Teacher Mindset for This Unit

This is the unit that most requires the teacher to model ethical reasoning — not just teach it.

Teacher Guide · Unit 5 · Package A

Differentiation by Cognitive Level

All students engage with the same essential question, the same Five Faces and the same Ethical AI Pledge. Differentiation adjusts cognitive demand and scaffolding depth — not the conceptual bar.

5.1 · Grade-Band Adaptation Matrix

Grades 7–8

Remember & Understand

Recognise → Match → Commit

Grades 9–10

Apply & Analyse

Analyse → Trace Impact → Propose

Grades 11–12

Evaluate & Create

Adjudicate → Draft Policy → Advocate

Activity	🟦 Grades 7–8	🟪 Grades 9–10	🟩 Grades 11–12
A1 · Capstone Bridge	Recall the four prior units in one word each; which could hurt someone if misused?	Map each prior competency to a specific harm.	Is ethical AI use individual or systemic — and what is a personal pledge worth either way?
A2 · AI Harm Mapping	Pre-selected clearest-narrative card (A or E); sentence starters provided.	Any case card; full Harm Analysis, all three sticky-note types.	Research a real, recent case not on the card list; produce a 1-page Harm Analysis.
A3 · Five Faces of Risk	Visual reference card; match the case to a Face.	Identify which Face(s) apply and explain why.	Propose a Sixth Face; define, exemplify and justify it.
A4 · Stakeholder Analysis	Pre-filled stakeholder table; add what each could have done differently.	Full analysis: all five groups and their possible actions.	Draft an advocacy statement — one systemic change, and who it should target.
A6 · Policy Audit	Simplified policy extract; 3 of the 5 Audit Questions.	Full 5-question audit on a policy of their choice.	Draft a school AI policy: at least 3 scenarios and 3 accountability mechanisms.
A7 · The Five Lenses	Match each lens to a given scenario from a provided list.	Name a default lens and a challenge lens, each with a personal example.	Apply all five to one real AI use, then resolve a tension between two of them explicitly.
A8 · Ethical AI Pledge	Sentence scaffold for Sections 1–5.	Full Pledge, all five sections, individual and unscaffolded.	Extended Pledge with an advocacy commitment; presented and defended publicly.
D6-C · Creator's Defence	Two questions instead of three; the question bank is shared in advance.	Three unseen questions drawn at random from the bank.	Three unseen questions plus one follow-up challenge from the student panel.

5.2 · Within-Band Support Levels

The grade band sets the cognitive demand; the support level sets the scaffolding. An advanced Grade 7 student receives the Grade 7 task with an extension — never the Grade 11 task.

Level	Harm Mapping (A2)	Five Faces (A3)	Policy Audit (A6)	The Pledge (A8)
🟡 Foundation	Pre-selected Case Card (A or E); structured Harm Analysis worksheet with sentence starters	Visual reference card; match the case to a Face rather than identify it from scratch	Simplified policy extract, key sentences only; 3 of the 5 Audit Questions	Sentence-scaffold template for Sections 1–5
🔵 Core	Group selects any Case Card; full Harm Analysis with all three sticky-note types	Identify which Face(s) apply and explain the reasoning	Full 5-question audit on a policy of their choice	Full Pledge, all five sections, individual and not scaffolded
🟢 Advanced	Research and present a real, recent AI harm case; produce a 1-page Harm Analysis	Propose a Sixth Face: define it, exemplify it, and justify its own category	Draft a proposed AI use policy for the school: 3 scenarios, 3 accountability mechanisms	Extended Pledge including an advocacy commitment beyond personal behaviour

5.3 · No-Device / Low-Tech Fallback

Unit 5 is device-independent by designCase cards, stakeholder tables, policy extracts and the Pledge are all paper-based. Only the Grades 11–12 research extensions in A2 and A6 need access — supply printed case summaries and printed policy extracts instead. The cognitive demand is unchanged.

Differentiation Reflection

Which grade band did you teach today? What did you adjust on the spot that this matrix did not anticipate?

Teacher Guide · Unit 5 · End

AI Explorer