📚 Unit Plans → 🌍 AI Ethics Learning Resources → 📄 AI Ethics and Bias
Years 8–13 Hangarau / Technology

Unit 7 Bias and Fairness • Years 8–13 • Inquiry worksheet • Print-ready

AI Ethics and Bias

Use this handout to help ākonga test AI outputs rather than simply consume them. Students compare responses, identify bias or omission, and decide what a fairer, safer, and more culturally responsible answer would require.

Ingoa / Name
Akomanga / Class

Best for

Lesson 2 or 3 inquiry, fairness checks, debate preparation, and any task where students must analyse the quality of an AI response rather than just use it.

Kaiako use

Bring one or two AI outputs, prompts, or screenshots to analyse together. The task is strongest when students can compare what the tool includes, flattens, or leaves out.

Ākonga use

Students can compare outputs, identify who may be advantaged or harmed, and build a short recommendation about what safer AI use should look like.

Free fairness check, premium local adaptation

Use this worksheet as written. For local or iwi-specific analysis, begin with material supplied or approved by the relevant authority. Te Wānanga may help format a source-bound worksheet; it cannot generate the cultural authority or local position.

  • Swap in local school decisions, community issues, or current news outputs.
  • Generate a junior version with sentence starters or a senior version with deeper critique.
  • Turn the final recommendation into a report, speech, or evaluation paragraph.
Adapt this in Te Wānanga Open Creation Studio Compare teacher plans

Kaiako planning snapshot

  • Use length: 25-40 minutes for one shared output, longer if students compare multiple examples.
  • Grouping: Pairs or small groups work best so students challenge each other’s reading of the output.
  • Prep: Pre-select one or two outputs that contain enough substance to analyse. For Māori-data or cultural claims, also supply a named source from the relevant Māori authority; Te Mana Raraunga’s principles are a suitable national starting point for Māori data sovereignty.
  • Differentiation: Support learners can complete the quick fairness check; extension learners can compare two outputs and write a stronger recommendation.
  • Teaching move: Keep asking “who benefits, who is erased, and what would a safer answer include?”
Fairness Bias Critical literacy

Resources already provided

  • Quick fairness check and comparison table
  • Risk, harm, and benefit prompts
  • Māori data sovereignty lens questions
  • Recommendation-writing scaffold

What to print: one copy per student or pair plus one teacher-selected AI output, prompt, or screenshot to analyse.

Ngā Whāinga Akoranga / Learning Intentions

  • We are learning how to test AI outputs for fairness, bias, and missing perspectives.
  • We are learning how digital tools can reproduce unequal power relationships.
  • We are learning to check whether a cultural or data-authority claim is supported, attributed, and kept within the scope of a named source.

Paearu Angitu / Success Criteria

  • I can identify at least one bias, omission, or risk in an AI output.
  • I can explain who may be advantaged or harmed by that response.
  • I can suggest what a fairer or safer response would need to include.

Why this matters in Aotearoa

AI outputs can sound fluent while misrepresenting people, flattening local context, or using Māori knowledge without clear provenance. Good critique identifies the exact claim, checks it against a named source, and records what the source does and does not establish. Ākonga do not decide whether an answer is culturally safe, tika, or mana-enhancing for other people.

1. Quick fairness check

  • Whose voice is centred in the answer?
  • Who or what is missing?
  • What assumptions does the AI seem to treat as normal?
  • What harm could follow if someone trusted this answer completely?

2. Compare one output closely

What the AI does well Bias / omission / risk How the answer should improve

3. Māori data sovereignty — source check

Te Mana Raraunga’s principles state that Māori have authority over Māori data and that all data has whakapapa: its provenance, purpose, context, and parties should be visible. Use that source, or a more relevant named iwi or hapū source, to answer these questions.

What data or knowledge is involved, and what provenance does the source establish?

Who does the source identify as holding authority over its collection, interpretation, use, or reuse?

What consent, access, accountability, or future-use limits does the source require?

Which parts of the AI output are unsupported or outside the source’s scope?

4. Recommendation scaffold

My judgement about this tested output

State the exact claim or omission, cite the named source, identify who may be affected, and limit your recommendation to the output and task you analysed.

Teach this tomorrow

Print / share

  • This worksheet
  • One teacher-selected AI output or screenshot

Decide before class

  • Whether students analyse one shared output or compare two
  • Whether the final response is discussion-based or written

Look for by the end

  • Students can identify one concrete bias or omission
  • Students can explain what a stronger response would need to change

What to do next

Use this worksheet to sharpen fairness judgement, then move into the AI Bias Detection Lab or a class debate built in Te Wānanga or Creation Studio.

Open in Te Wānanga Return to My Kete

Aronga Mātauranga Māori

This activity uses Te Mana Raraunga’s published principles rather than a generic Māori ethics scale. Those principles locate authority with Māori, describe data as carrying whakapapa, and require attention to provenance, consent, accountability, benefit, access, and kaitiakitanga. Ākonga can check an AI output against those stated principles; they cannot speak for a community or infer whether that community would trust the answer.

Ngā Rauemi Tautoko · Support Materials