🔍 Inclusion Case Studies — AI in Real Classrooms

Educators across Aotearoa are already adapting AI to uplift identity, access, and inclusion.

Here are three grounded scenarios showing how inclusive AI teaching works in practice — and what to avoid.

Case Study 1: ESOL Learners — Preserving Voice While Building Confidence

Context

An ESOL learner uses AI to improve grammar and vocabulary. Their writing becomes highly fluent — but no longer sounds like them.

The Challenge

How can AI support language development without erasing identity?

What the Educator Did

• Allowed AI for grammar and sentence clarity

• Required a “voice reflection”:

“What did AI change? Which changes sound like you, and which don’t?”

• Added a 1-minute audio explanation in the learner’s own voice

• Reassured learners that imperfection is part of authentic communication

Outcome

Confidence increased.

Grammar improved.

Identity remained visible.

AI became a language scaffold — not a ventriloquist.

🪶 Kaupapa Māori Insight

Mana motuhake matters.

Support should enhance a learner’s voice, not replace it.

Case Study 2: Neurodivergent Learners — Reducing Cognitive Load

Context

A neurodivergent learner struggles with multi-step instructions and organisation.

The Challenge

How can AI help reduce overwhelm while still supporting independent learning?

What the Educator Did

• Gave permission to use AI to:

break tasks into steps

create personalised study plans

turn instructions into plain language

• Required a short check-in:

“Which parts did AI help make clearer?”

• Ensured AI was not used for the final answers

Outcome

Reduced cognitive load → increased engagement.

Learner gained confidence in managing workload.

🪶 Kaupapa Māori Insight

This is manaakitanga in practice — care that removes barriers without removing agency.

Case Study 3: Māori & Pacific Learners — Locally Grounded Examples

Context

A business class uses AI to generate marketing examples.

The outputs are generic and culturally off-key.

The Challenge

How do we ensure AI supports culturally grounded learning, not generic global norms?

What the Educator Did

• Asked students to critique the AI output:

“What’s missing? What’s inaccurate? Whose worldview is centred?”

• Required learners to rewrite examples using:

local context

whānau and community networks

language and values meaningful in the region

• Created an optional “teach the AI” activity:

“What would the AI need to know about your community to make this example accurate?”

Outcome

Learners saw AI’s cultural gaps instantly.

Their rewrites were richer, more grounded, and more confident.

🪶 Kaupapa Māori Insight

This is whakapapa in action — tracing the origins of knowledge and restoring what AI leaves out.

Case Study 4: Disability Support — Assistive AI Without Stigma

Context

A learner with dysgraphia uses AI for note-taking and summarising.

The Challenge

Ensuring assistive AI use is:

• normalised

• transparent

• not viewed as “cheating”

What the Educator Did

• Made assistive AI tools optional for all learners

• Normalised diverse tools in class demonstrations

• Encouraged students to reflect on how AI supported accessibility

Outcome

AI became a universal design tool — not a special exception.

🪶 Kaupapa Māori Insight

Whanaungatanga: shared tools build shared belonging.

Case Study 5: Rural Learning — Overcoming Resource Gaps

Context

A rural class with limited access to specialists uses AI to support STEM explanations.

The Challenge

Ensuring students don’t treat AI as the only source of truth.

What the Educator Did

• Taught students how to cross-check AI explanations

• Used AI as a concept starter, not an authority

• Had students interview whānau or local experts as counterpoints

Outcome

Learners saw AI as one voice among many — not the oracle.

🪶 Kaupapa Māori Insight

Kaitiakitanga means using tools wisely and grounding them in local knowledge systems.

Case Study 6: Large Classes — Personalisation at Scale

Context

A tertiary provider uses AI to give formative feedback in a 120-student class.

The Challenge

Avoiding generic AI feedback that feels impersonal or biased.

What the Educator Did

• Combined AI-generated feedback with quick human audio notes

• Required a student reply:

“Which feedback will you act on first, and why?”

• Ensured feedback addressed:

strengths

structure

examples

cultural relevance

Outcome

Human touch preserved.

AI workload reduced.

Learners felt seen.

🪶 Kaupapa Māori Insight

Feedback is relational — a form of ako.

AI can support it, but cannot replace reciprocal learning.