Teacher Discussion Guide: AI Bias, Racism & Technology

Your Role: Create brave space where students can explore difficult truths about technology and society while feeling culturally safe and supported.

Challenge 1: "AI isn't racist, it's just a computer"

Why This Comes Up: Students may resist the idea that technology can be biased, seeing computers as objective.

Facilitation Response:

"You're right that computers don't have racist intentions. But AI learns from data created by humans, and if that data reflects human biases and inequalities, the AI will learn and amplify those patterns. Let me show you a concrete example..."

Follow-Up Activity: Show the MIT facial recognition study (34.7% error for dark-skinned women vs 0.8% for light-skinned men). Ask: "This AI doesn't hate anyone, so why does it work worse for certain groups?"

Challenge 2: Defensive Reactions from Pākehā Students

Why This Comes Up: When discussing how AI bias harms Māori/Indigenous peoples, some Pākehā students may feel personally accused or become defensive.

Facilitation Response:

"We're not saying anyone in this room is racist. We're examining how systems - including AI systems - can produce unequal outcomes even when no individual intends harm. Understanding this helps us build better systems that work fairly for everyone."

Redirect Focus: From "who's to blame?" to "how can we fix this together?" Emphasize students' role as future designers and advocates.

Challenge 3: Māori Students Bearing Burden of Education

Why This Comes Up: Māori students may be asked to explain or validate Māori perspectives, putting emotional labor on those most affected by AI bias.

Prevention Strategy:

• Don't put Māori students on the spot to "speak for all Māori"
• Provide resources (readings, videos) where Māori voices are already documented
• Create space for Māori students to share IF they want to, never as requirement
• Acknowledge: "Māori students shouldn't have to educate the class - that's my job as teacher"

Challenge 4: "Why Focus on Māori? Other Groups Face Bias Too"

Why This Comes Up: Attempts to deflect from specific Māori/Indigenous issues by generalizing to all marginalized groups.

Facilitation Response:

"Absolutely - AI bias affects many groups including women, disabled people, LGBTQ+ communities, and other ethnic minorities. In Aotearoa, we have particular responsibility to understand how AI affects tangata whenua. Understanding Māori experiences with AI bias helps us recognize patterns that affect other groups too. We can address multiple forms of bias - it's not either/or."

1. Establish Ground Rules Early

Before Lesson 2, set clear expectations:

• Listen to learn: We're exploring complex ideas together
• It's okay to be uncomfortable: Growth happens outside comfort zones
• Ask genuine questions: No question is "stupid" if asked with genuine curiosity
• Respect lived experience: People affected by bias are experts on their own experiences
• Focus on systems, not individuals: We critique systems, not people in the room

2. Use Data & Evidence

Ground discussions in documented facts, not opinions:

• Reference specific research (MIT study, Amazon hiring AI, etc.)
• Show actual AI outputs demonstrating bias
• Use numbers and statistics to make bias concrete
• Cite Indigenous data sovereignty experts (Te Mana Raraunga)

3. Center Indigenous Voices (Not Just You)

Let Māori and Indigenous experts speak for themselves:

• Play video clips of Te Hiku Media team explaining Papa Reo
• Share quotes from Te Mana Raraunga reports
• Use case studies where Indigenous communities describe their own experiences
• Invite local Māori tech professionals as guest speakers if possible

4. Validate Emotions, Redirect to Learning

When emotions run high:

• Acknowledge: "I can see this topic brings up strong feelings. That's actually appropriate - we should care about justice."
• Pause if needed: "Let's take a breath. This is important and we want to discuss it thoughtfully."
• Refocus on learning: "What can we learn from this reaction? What does it tell us about how we think about technology and fairness?"

Q: "Isn't focusing on bias just being too sensitive? Can't we just use the technology?"

A: "When facial recognition misidentifies you 35% of the time but works perfectly for others, that's not sensitivity - that's a broken tool. When hiring AI automatically rejects qualified candidates because of their names or backgrounds, that's not political correctness - that's discrimination. We're learning to recognize when technology is actually failing to serve everyone equally. That's not being sensitive, it's being smart and fair."

Q: "Why should we care about Māori data sovereignty? Isn't all data the same?"

A: "Imagine your family's private photos, your whakapapa, your health information being used to train AI without your permission or knowledge. Now imagine that happening to an entire culture's knowledge systems. That's what's at stake. Te Hiku Media shows a different path - Māori controlling Māori data, deciding how it's used, benefiting from the AI they helped create. That's not special treatment, it's basic fairness and respect."

Q: "Can't we just make AI 'colorblind' to avoid bias?"

A: "Ignoring race or culture doesn't eliminate bias - it can make it worse by preventing us from seeing and addressing disparities. Better approach: Include diverse perspectives in AI development, test for bias across different groups, and design systems that actively work toward equity. Being aware of differences allows us to ensure fair treatment, not discriminatory treatment."

If a Student Makes a Racist Comment:

1. Interrupt immediately: "Hold on. That statement is harmful because..."
2. Name the harm: Be specific about why it's problematic
3. Restate class values: "In this class, we treat all cultures and people with respect"
4. Redirect learning: "Let's examine why that belief is inaccurate..."
5. Follow up privately: One-on-one conversation with student after class
6. Support affected students: Check in with Māori students who may have been hurt

If Discussion Becomes Heated Argument:

1. Pause the discussion: "Let's pause for a moment. This is clearly important to people."
2. Acknowledge emotions: "Strong feelings show we care about justice. Let's channel that energy productively."
3. Refocus on evidence: "Let's return to what the research shows..."
4. Structure the conversation: "One person speaks at a time. We listen to understand, not to argue."
5. Take a break if needed: "We'll come back to this tomorrow after we've all had time to think."

If You Don't Know the Answer:

It's okay to say: "That's a great question and I don't have a complete answer. Let's research this together and come back to it next lesson."

Or: "I'm not the expert on Māori perspectives here. Let's find voices from Māori tech professionals or Te Mana Raraunga to learn from."

Never: Make up answers or speak for Māori perspectives you don't have authority to represent.

Cultural Safety Practices:

• Check in privately: "How are you feeling about Unit 7 content? Is there anything you need from me?"
• Offer choice: "You can share your perspective if you want, but you're never required to"
• Validate experiences: "Your lived experience of technology bias is valuable knowledge"
• Provide alternatives: If a lesson feels too personal/painful, offer alternative assignment approaches
• Celebrate strengths: Highlight Indigenous tech excellence (Te Hiku Media, Te Mana Raraunga)