When people talk about Māori and AI in the same sentence, the kōrero usually starts with risk. Bias in the models. Data taken without consent. Language flattened by systems trained on someone else’s world. Those risks are real, and naming them matters.
But that is not where the story ends, and honestly, it is not where my head is at. When I look at what is coming, I feel optimistic — not naively, but deliberately. Here is why.
Because Māori are already leading, not catching up
The best answer to "will AI serve Māori?" is that Māori are already building AI that does. Te Hiku Media trained world-class speech recognition for te reo Māori on decades of archival recordings — and kept governance of that data with the communities the language belongs to. Researchers are formalising Māori algorithmic sovereignty so that values like whakapapa and mana shape how systems are designed, not just how they are criticised afterwards.
That precedent changes everything. It proves the choice is not "adopt someone else’s AI or opt out". There is a third path: build it ourselves, on our terms.
Because the barriers to building have never been lower
This is the part I feel personally. I am a mum of four in Taumarunui building a games and learning company without investors, without a computer science degree, and without asking permission. The same AI that worries people is one of the reasons a pakihi like mine can exist at all — it lowers the cost of creating, designing, coding, and learning as you go.
Now multiply that. Every marae, kura, and whānau business has people with ideas that never had a pathway to becoming real. When the tools get cheaper and more capable, those ideas start shipping. Some of them will be learning tools our tamariki actually recognise themselves in.
Because our ways of learning fit where education is heading
Education is slowly moving toward things Te Ao Māori never left behind: learning through story, through doing, through relationships, across generations. Ako — the idea that teaching and learning flow both ways — sounds a lot like where the best modern learning design is trying to get to. We are not adapting to the future of learning. In some ways, it is catching up to us.
Optimism with open eyes
None of this makes the risks disappear. Data sovereignty needs defending. Bias needs auditing. Access needs to reach every whānau, not just the connected ones. Optimism is not the absence of vigilance — it is the belief that showing up and building is worth it.
The future of AI and learning in Aotearoa is not something that happens to Māori. It is something Māori are already making.