AI Ethics Design Project Rubric
Unit 7: Digital Tech & AI Ethics | Years 9-10 | 100 points total
Project Task:
Design an AI tool or feature guided by te ao Māori values (tikanga). Create: (1) Prototype/mockup, (2) Ethical framework document, (3) 5-minute presentation explaining how your design prioritizes people over profit and prevents harm.
Deliverables: Visual prototype, 500-word ethics document, presentation slides, demo/explanation.
| Criterion | Excellent (A) 20pts | Proficient (B) 16-19pts | Developing (C) 11-15pts | Beginning (D-E) 0-10pts |
|---|---|---|---|---|
| Tikanga Integration (20 pts) | Design deeply grounded in specific Māori values (manaakitanga, kaitiakitanga, whanaungatanga, rangatiratanga). Each design choice explicitly justified by tikanga. Shows nuanced understanding of how values conflict with typical AI development. | Design incorporates 2-3 Māori values. Most choices linked to tikanga. Shows understanding of value conflicts. | Mentions Māori values but integration is superficial. Limited explanation of how values shaped design. Unclear on conflicts. | No clear tikanga integration or only surface-level mention. Design could exist without Māori values. |
| Problem-Solution Fit (15 pts) | Identifies real, specific problem affecting communities. Solution directly addresses root cause (not just symptoms). Explains why AI is appropriate tool. Evidence problem exists. | Identifies clear problem. Solution addresses it. Some evidence. AI choice justified. | Problem vague or not well-researched. Solution doesn't fully address it. Weak justification for using AI. | No clear problem identified. Solution seems disconnected. No justification for AI approach. |
| Harm Prevention & Safety (20 pts) | Identifies 3+ potential harms (bias, privacy, exploitation). Specific safeguards for each. Considers vulnerable populations. Transparent about limitations. Accountability mechanisms clear. | Identifies 2-3 harms with safeguards. Some consideration of vulnerable groups. Basic accountability. | Lists 1-2 harms but weak safeguards. Limited consideration of who might be harmed. Vague accountability. | No harm analysis or only surface-level. No safeguards. No accountability plan. |
| People Over Profit (15 pts) | Design explicitly rejects profit-first logic. Shows how success is measured by community benefit, not revenue. Governance structure prevents corporate capture. Clear about who controls the AI. | Design prioritizes wellbeing. Non-profit focused. Some governance clarity. | Claims to prioritize people but design allows profit motives. Governance unclear. | No clear prioritization of people. Could easily be commercialized. No governance plan. |
| Prototype Quality (15 pts) | Visual prototype is clear, polished, and demonstrates key features. User interface thoughtfully designed. Accessible (considers disabilities, digital literacy). Includes enough detail to understand functionality. | Clear prototype showing main features. Decent UI. Some accessibility consideration. Functional. | Rough prototype. Hard to understand features. Little UI thought. No accessibility consideration. | Minimal or no prototype. Can't visualize how it would work. |
| Presentation & Communication (15 pts) | Confident, engaging 5-min presentation. Clear explanation of problem, solution, tikanga integration, harm prevention. Good slides. Answers questions thoughtfully. Stays on time. | Clear presentation covering all elements. Decent slides. Answers most questions. Near time limit. | Unclear presentation missing elements. Poor slides. Struggles with questions. Time issues. | Minimal presentation. Can't explain design. No slides or very poor quality. Significantly over/under time. |
Grade Conversion:
- A (Excellence): 90-100 points
- B (Merit): 75-89 points
- C (Achieved): 50-74 points
- D-E (Not Achieved): 0-49 points
📚 Teacher Marking Guide:
Look For Excellence:
- Student names specific tikanga principles and traces design choices back to them
- Student identifies who is vulnerable to harm and designs specific protections
- Student questions why typical AI prioritizes profit and designs alternative
- Student considers governance: who controls the AI? who benefits? who decides?
- Prototype is thoughtful, not just "slap Māori values on existing tech"
Red Flags:
- Māori values mentioned but not integrated (tokenistic)
- No analysis of potential harms or very surface-level
- Design easily commercializable (profit motives still dominant)
- No consideration of who controls the technology
- Prototype is just existing tech with Māori aesthetic
Example Projects (A-Level):
- Te Reo Chatbot: Prioritizes language revitalization over commercial use. Kaitiaki license model (Te Hiku Media-inspired). Free for educational use, requires ethical consent for commercial. Governance by rūnanga.
- Mental Health AI: Connects youth to community support, not just symptom management. Designed with hauora (holistic wellbeing) - addresses social isolation, not just individual "disorder." Transparent about data use, locally hosted.
- Climate Monitoring Tool: Combines mātauranga Māori environmental indicators with sensor data. Empowers iwi to protect their rohe. Data sovereignty: iwi control all data, decide what's shared. Builds whanaungatanga between kaitiaki.