AI Ethics in Practice
Ākonga deliberate over a real decision using competing ethical frameworks, Māori data sovereignty, and community consequences. The emphasis is on explaining and revising reasoning before choosing an individual or collective action.
- Scaffold: Hui protocol, stakeholder role cards, and an evidence-to-values reasoning frame
- Seminar outcome: Defensible ethics framework and a specific community-action commitment
🌅 Opening & Decision Protocol
🎥 Media Anchor
Video: "The Coded Gaze: Unmasking Algorithmic Bias" — Joy Buolamwini
- How can automated systems amplify inequity if left unchecked?
- What evidence should be required before deploying high-impact AI?
Decision protocol: evidence, interpretation, uncertainty, authority.
Opening Protocol (5 minutes)
- Evidence: Identify the source-backed facts in the case.
- Interpretation: State the ethical principle used and how it changes the judgement.
- Limits: Record uncertainty and the person or group holding the next decision.
🎯 Learning Objectives & Success Criteria
By the end of this lesson, ākonga will be able to:
- Construct: Build an AI ethics framework from named sources and explicit ethical principles
- Apply: Use ethical frameworks to evaluate real-world AI scenarios
- Distinguish: Separate published evidence, ethical interpretation, uncertainty, and decisions held by affected authorities
- Plan: Design a permission-aware next step for promoting ethical AI
Success Criteria — Ākonga will demonstrate:
- ✓ Clear articulation of personal AI ethics principles
- ✓ Practical application of ethics frameworks to complex scenarios
- ✓ Accurate use of named sources, including scope and authority limits
- ✓ Realistic next steps that do not presume community permission
Phase 1: Ethics Framework Construction — Building Your AI Ethics Compass (25 minutes)
Personal AI Ethics Framework Workshop
20 minutes construction + 5 minutes sharingStep 1: Core Values Foundation (8 minutes)
Working individually, students identify their core values for AI ethics by completing this reflection:
Published Māori data-sovereignty constraints
Using Te Mana Raraunga’s published principles, choose two or three questions that matter for this AI case. Keep the source attached; these are not personal definitions of mātauranga Māori.
- ☐ Authority: Who may decide how the data is collected, interpreted, used, or reused?
- ☐ Provenance: Are purpose, context, source, and involved parties visible?
- ☐ Consent: What permission exists, and can it be withdrawn?
- ☐ Accountability: Who answers for harm, misuse, or error?
- ☐ Access: Who can inspect, correct, or remove data?
- ☐ Benefit: Who receives the benefits created from the data?
- ☐ Kaitiakitanga: What care and future-use duties does the source require?
- ☐ Scope: What does this source not authorise us to conclude?
Universal Ethical Principles Selection
Choose two or three additional ethical principles and keep their definitions explicit:
- ☐ Human Dignity — Respect for inherent worth of all people
- ☐ Fairness — Equal treatment and just distribution of benefits/risks
- ☐ Transparency — Openness about how AI systems work
- ☐ Privacy — Protection of personal information and autonomy
- ☐ Accountability — Clear responsibility for AI decisions and impacts
- ☐ Beneficence — Doing good and maximising positive impact
- ☐ Non-maleficence — Avoiding harm and minimising negative impacts
- ☐ Sustainability — Long-term environmental and social responsibility
Step 2: Framework Integration (7 minutes)
Create your integrated AI ethics framework by completing these statements:
My Core Principle:
"When evaluating AI systems, the most important thing to me is..."
Student writes their overarching principle combining chosen values
My Decision Process:
"When facing an AI ethics dilemma, I will ask these questions..."
Student creates 3–4 key questions based on their chosen values
My Red Lines:
"AI systems should never..."
Student identifies 2–3 absolute boundaries based on their values
Step 3: Framework Testing (5 minutes)
Quickly test your framework on this scenario:
Quick Test Scenario: A social media platform uses AI to curate content feeds. The AI increases engagement but also tends to show more extreme or polarizing content because it gets stronger reactions. Using your framework, is this ethical? What questions would you ask?
Framework Sharing (5 minutes): Students pair up and share their frameworks, discussing similarities and differences.
Phase 2: Real-World Application — Ethics Frameworks in Action (25 minutes)
AI Ethics Decision-Making Lab
Complex Scenario Analysis (20 minutes):
Working in groups of 3–4, apply your individual frameworks to these fictional scenarios. They are prompts, not evidence that the described events occurred. For each one, list the factual claims that would require sources and the decisions that would require affected-community authority.
Scenario 1: AI-Powered Healthcare in Rural Communities
Fictional context: A health provider proposes an AI diagnostic tool for a named rural service area. Its published evaluation does not yet establish equivalent performance for the Māori patients it would serve.
The Dilemma:
- The AI could provide life-saving early detection of diseases
- But it may be less accurate for Māori patients due to training data bias
- The provider reports a need for better specialist access
- But using biased AI could worsen health inequities
- Waiting for better AI could delay help for years
Additional Factors:
- The affected health authority and patient representatives must shape the decision
- Any relationship with rongoā or other health practice requires the relevant practitioners’ authority
- Data sovereignty concerns about health information
- Cost and accessibility issues
Group Analysis Process:
- Each member applies their personal framework (5 minutes)
- Share individual recommendations and reasoning (5 minutes)
- Find common ground and build group consensus (5 minutes)
- Develop specific action recommendations (5 minutes)
Scenario 2: AI Content Moderation and Cultural Expression
Fictional context: A platform’s moderation system flags public posts from a named Māori cultural organisation. The organisation says some removals misread the context. No claim about the scale or cause is established until the case log and appeal records are examined.
The Dilemma:
- Automated moderation is proposed to handle a very large volume of posts
- But it's censoring legitimate cultural expression
- Manual review by humans is too slow and expensive
- But AI lacks cultural context and nuance
- Platform wants to be culturally respectful but also safe
Additional Factors:
- Different cultures have different norms about appropriate content
- The named organisation uses the platform to share public material
- Advertisers pressure platform to be "brand-safe"
- Free speech and cultural preservation concerns
Scenario 3: AI in Education and Cultural Knowledge
Fictional context: A school trials an AI tutoring system. A teacher audit finds unsupported Māori-related claims and a narrow source base. The school must decide what the tool may do, what requires verified curriculum sources, and what must remain with qualified people.
The Dilemma:
- AI could democratize access to personalized education
- But fluent output may misrepresent knowledge or exceed its sources
- Students need help outside school hours
- But AI may create homogenized, culturally narrow thinking
- Teachers want support but require evidence, attribution, and qualified review
Additional Factors:
- Whānau and relevant Māori authorities must be able to shape any local cultural use
- The vendor has not demonstrated authority over Māori language or knowledge
- Questions about who owns and controls educational AI
- Concern about replacing human teachers and relationships
Group Decision Presentation (5 minutes):
Each group presents their scenario decision in 90 seconds, covering:
- What named sources and ethical frameworks guided their decision
- Their recommended course of action
- How they balanced competing values and interests
- What remains uncertain and who holds the next decision
Phase 3: Evidence-to-Action Planning (20 minutes)
Permission-Aware Next-Step Workshop
Evidence and permission plan (15 minutes):
Working individually, plan one bounded next step. Use a named public source, identify the people affected, and do not contact, publish about, or organise on behalf of a community without the appropriate permission.
Step 1: Issue Focus (3 minutes)
Choose one AI ethics issue established by a named source:
- ☐ AI bias affecting local hiring or services
- ☐ Social media algorithms and youth mental health
- ☐ AI in local healthcare or education
- ☐ Surveillance technology in public spaces
- ☐ AI impact on local jobs or economy
- ☐ Language or cultural-data governance in a named public project
Step 2: Stakeholder Mapping (4 minutes)
Identify key people and groups who need to be involved:
Decision Makers:
Who has power to make changes?
- Local government officials
- School administrators
- Business leaders
Affected people:
Who does the source identify as affected, and who is missing?
- People named in the source
- People represented in the system’s data
- People who may bear an untested risk
Allies and Advocates:
Who might support this cause?
- Teachers and educators
- Community organisations
- Local media
Step 3: Action Strategy (5 minutes)
Plan specific actions you could realistically take:
Evidence preparation:
- ☐ Research the issue more deeply
- ☐ Record what the source establishes and what it does not
- ☐ Identify the authority or organisation relevant to the next decision
- ☐ Draft questions without sending or publishing them
Permission and review:
- ☐ Ask the kaiako which contact or audience is appropriate
- ☐ Identify consent and attribution requirements
- ☐ Seek qualified review of any te reo Māori or cultural material
- ☐ Revise the proposal before any external use
Possible next step after approval:
- ☐ Submit a sourced question or recommendation to the approved audience
- ☐ Record the response accurately
- ☐ Revise the claim, design, or safeguard
- ☐ Stop if permission or authority is absent
Step 4: Success Metrics (3 minutes)
How will you know whether the plan is responsible and evidence-bound?
- Every factual claim has a named source
- The affected people and decision-maker are not assumed
- Permission, attribution, review, and stop conditions are visible
- The proposed action remains within the student’s role
Internal design critique (5 minutes):
Students share plans inside the class and check one another’s evidence, scope, permission, authority, and stop conditions. This critique does not authorise external contact or publication.
🌅 Whakamutunga — Reflection & Closing
Ethical Leadership Commitment (5 minutes)
Personal Ethics Integration Reflection:
Students complete individual reflection:
- How has building your personal AI ethics framework changed your thinking about technology?
- What was most challenging about applying ethics frameworks to real-world scenarios?
- What did the named Māori data-sovereignty source require, and what remained outside its scope?
- What is one permission-aware next step you could take after kaiako review?
Closing Circle — Ethical Leadership Commitment:
Students share one word describing their commitment to ethical AI leadership, followed by teacher reflection:
"Mā te huruhuru ka rere ai te manu"
Just as birds need feathers to fly, we need ethical frameworks to navigate the complex world of AI. Your framework is your compass — use it to make decisions that honour both innovation and justice. The future of AI depends on ethical leaders like you taking action in your communities.
📊 Assessment & Next Steps
Formative Assessment — Today's Evidence:
- Framework Construction: Depth and integration of personal AI ethics framework
- Applied Analysis: Quality of ethical reasoning in complex scenarios
- Source and authority: Accurate source use, visible scope, and decisions reserved to affected authorities
- Action planning: Realistic evidence, permission, review, and stop conditions
Preparation for Lesson 4:
- Framework Testing: Use your ethics framework to evaluate one AI tool you use regularly
- System scan: Identify one AI system and a named source documenting its purpose or effects
- Plan review: Revise one next step after checking evidence, permission, and authority with the kaiako
🛠️ Teacher Resources & Adaptations
Ethics Framework Resources:
- Stanford Encyclopedia of Philosophy: Comprehensive ethics theory background
- AI Ethics Lab: Practical frameworks for AI ethics education
- Te Mana Raraunga: Published Māori data-sovereignty principles
- Future of Humanity Institute: Applied ethics in emerging technology
Scenario Development Support:
- Local News Sources: Current AI implementation in your region
- Named affected organisations: Public statements about specific AI impacts and concerns
- Ethics Bowl Resources: Additional complex scenarios for practice
- Case Study Databases: Real-world AI ethics dilemmas
Differentiation Strategies:
- Framework Complexity: Simplify or expand ethical framework elements
- Scenario Selection: Choose scenarios matching student interests and experiences
- Action planning: Scale evidence and proposal work to student capacity; external action requires approval
- Source boundaries: Supply named sources and keep local cultural decisions with the relevant authority
Kaiako Planning Snapshot
Digital Technologies / Social Studies — AI Ethics & Te Ao Māori — Years 9–10
Learning context: Digital Technologies and ethical reasoning through a te ao Māori lens (Years 9–10).
Ngā Whāinga Akoranga — Learning Intentions
- Explain how AI systems work at a conceptual level and identify their ethical dimensions
- Analyse AI bias and algorithmic justice through a te ao Māori and tikanga framework
- Evaluate the implications of AI for tino rangatiratanga, Māori data sovereignty, and digital futures
Differentiation & Proximal Guidance
Extension: investigate a real AI system for bias and propose a tikanga-aligned redesign. Scaffold: AI ethics checklist with guided prompts and vocabulary support. On-level: structured case study analysis. Entry: visual explainer of AI concepts with relatable everyday examples.
Audited curriculum connection
- Te Mātaiaho (2025) · Technology · Phase 4 (Years 9–10) · Knowledge: “Analysing AI applications in different sectors and predicting impacts on people and work”