A tool is not neutral

Picture this: A student asks a generative AI tool to “write an Indigenous creation story”. In return the AI produces a generic, pan-Indigenous narrative that is missing context, community and meaning. 

Generative AI tools are not culturally neutral. They reflect the dominant, Western-centric data they are trained on. 

This becomes especially problematic when students or teachers engage AI with Indigenous topics. 

Why does this matter when embedding First Peoples Principles of Learning?

The First Peoples Principles of Learning emphasize:

• holistic, experiential learning that is rooted in the land, identity, and relationships
• story, memory, and place as key ways of knowing
• respect for sacred knowledge and understanding that not all knowledge is for everyone to share
• learning is relational it is not transactional or generated on demand

In contrast, generative AI:

• pulls information from massive datasets – often without context, consent, or Indigenous authorship
• reflects dominant narratives, rather than diverse lived experiences
• risks treating cultural content as something that can be copied, rather than something sacred or community-held

Risks to be aware of

Cultural Misrepresentation: AI often creates pan-Indigenous or stereotypical responses. It lacks the nuance of nation-specific traditions, protocols, and language. It may also violate knowledge-sharing boundaries by generating sacred stories as public domain. 

Data Bias and Knowledge Gaps: Training data for the generative AI tools often reflect colonial worldviews and settler histories. Leading to Indigenous voices being underrepresented or tokenized. This can reinforce deficit narratives, such as trauma-focused content without context of strength and resilience. 

Misuse in Student Tasks: Students might ask generative AI to “teach them” about Indigenous culture. This inherently is bypassing authentic, community-based learning. There is also a copy-paste risk, where students may present the AI-generated content as a fact without checking its validity, or what it is missing. These surface-level “Indigenous perspectives” with no grounding in an actual community voice or meaning, leads to tokenism. 

Strategies to mitigate these risks

1. Critical framing and teacher guidance
   • Framing AI as a thinking partner, not a cultural authority
     • Encouraging students to compare the AI generated response to what they have already learned from Indigenous speakers, stories or sources
     • Encourage questioning: what is missing? whose knowledge is this?
2. Co-construct classroom norms
   • Set clear ethical guidelines
     • No generating sacred, spiritual, or traditional stories via AI
     • Use AI to support rather than replace relationship-based learning
3. Embed reflection through the FPPL lens
   • Does this AI generated output respect memory, story, and place?
   • Is this knowledge mine to share – or should I listen and learn instead?
   • How does this connect (or disconnect) from what I’ve learned in my community or on the land?

Better practices with including Indigenous content with AI

Using AI for skill-based tasks such as summarizing writing techniques, practicing reflective writing or revision, or brainstorming inquiry questions – not using it for content creation. 

Centre Indigenous voices first by using Indigenous sources such as books, guest speakers, or oral stories, as the foundation. Then use AI to support student reflections or to make cross-curricular connections. 

Use FPPL as the lens in which you analyze and critique generative AI outputs. Is this holistic? Is it relational? Does it reflect lived experience?

Centering respect over convenience

Generative AI can be a powerful classroom tool – but cultural respect, ethical reflection, and relational learning must come first, especially when working with Indigenous content. 

Educators have a responsibility to slow down, ask better questions, and listen before generating.

Practice using AI with care, curiosity, and cultural humility – and ensure your students are doing so too.