There’s a lot of talk about the need for data sovereignty. But what does that really mean when it comes to our reo and mātauranga Māori? Siena Yates has been talking to tech experts about the impact of “digital colonisation”.
Imagine if, when Pākehā first came to Aotearoa, Māori already had a heads-up about what was about to take place.
What if they knew exactly which strategies settlers would use to force a new way of life on them, and were able to develop the tools and opportunities to safeguard their whenua, mātauranga and reo? What if they were able to take what they wanted from te ao Pākehā and merge it with te ao Māori?
More importantly, what if, right now, there was a second wave of colonisation coming our way? What would we do differently?
It’s worth thinking about because, according to Te Taka Keegan, an artificial intelligence scholar, that’s exactly what’s happening.
Te Taka (Waikato-Maniapoto, Ngāti Porou, Ngāti Whakaue) is an expert in software engineering and AI. He’s co-director of the Māori AI Institute at Waikato University, where his mahi focuses on reo Māori technologies and Indigenous language interfaces.
He’s worked with Microsoft and Google to develop tools like the Microsoft Māori keyboard, Microsoft Office in Māori, Google search in Māori, and the Māori macroniser.
And he says we’re staring down the barrel of “digital colonisation”. He points to how Chat GPT is already speaking te reo Māori, composing waiata Māori and even writing karakia. It’s doing that by taking data it shouldn’t necessarily have access to.
According to Te Taka, we’ve all been “signing our lives away” on platforms like Facebook because they allow us to do things like make new connections, hold online classes, livestream hui, and stay in touch with our iwi, hapū, marae and whānau. But everything we put into that platform becomes data that can feed a tool like Chat GPT.
“When you sign up to Facebook or other social media platforms,” says Te Taka, “you’re prompted to agree to their terms, and you go, ‘Sure, whatever. How can this ever hurt me?’ But now, it’s starting to hurt us.”
He’s concerned that we’ll become too reliant on Chat GPT for reo growth and development. Every time we need a new kupu, phrase, waiata or karakia, we’ll turn to it as the easiest, most convenient option, rather than our original sources.
Ultimately, Te Taka fears that this will cause a shift in our reo. We’ll no longer have different dialects or learn kupu from different iwi. Our reo will become one homogenised language.
“That’s colonisation in terms of saying: ‘We’re all going to do it the same way — one way. It’s this way, and it’s a Pākehā way.’ It’s a loss of control and authority.”
It’s like one random American stopping at Auckland airport, listening to a few different conversations in the Koru Lounge, and then taking those snippets to the rest of the world, proclaiming: “This is the Māori language.”
And it’s not just the reo that Te Taka is worried about. “As Māori, we run with different stories, right? For us from Waikato and Maniapoto, Tāwhaki went up to the heavens to get the baskets of knowledge, but in other iwi, it was Tāne who went up. No one’s ever gonna get up and say: ‘You’re wrong’, because that’s the uniqueness and the beauty of being Māori. Being able to be different, have different stories. But Chat GPT says: ‘Nah, this is the one right answer.’ And that’s a colonising and very Pākehā perspective.”
So, what’s the solution?
“Why don’t we take these tools and cut them off from the mothership?” he says. He likens it to the ongoing debate around having bilingual road signs. It’s easy to make the sign, but who has the power to decide what’s on the sign?
“The technology’s not all that difficult. So, if it’s not working for us, let’s do it ourselves. Control our own story. That’s the attitude we need to have as Māori.”
When Chrissy Karena (Rangitāne o Tāmaki Nui a Rua, Te Āti Haunui) was growing up in Dannevirke, all the data on Māori pointed one way: down.
“All I ever got told through data was: ‘You’ll never be educated. You’re always going to be poor.’ Kids like us had no concept of the possibility that we could expect better,” says Chrissy. “Even though you don’t know it, you carry the burden of that data. It just settles within you.”
Even now, she’s inundated with negative data in her mahi as a health and cultural advisor at Te Whatu Ora. She refers to statistics that show Māori disproportionately affected by what seems like pretty much everything in the health system including diabetes, heart disease, cancer, obesity, poor oral health, respiratory issues and mental health.
What the data doesn’t show are the inequities and racism inherent in the health system that led to those results. It also doesn’t show how Māori health outcomes improve when iwi and kaupapa Māori are given support to do things their own way — as happened during Covid.
This is why, despite being 56 and, as she puts it, “kuia haere te roro” (her brain is ageing), Chrissy is studying at Poipoia, a data apprenticeship programme, in an effort to take back control.
“If we had the opportunity to gather data about ourselves, we wouldn’t necessarily gather the same information. We have our own ways of talking about and measuring wellbeing and how we see ourselves,” says Chrissy.
“That’s why I really see the need for Māori to work in the data space because if our data is being captured and analysed by our people, about and for our people, that’s us reclaiming our sovereignty and independence and designing our own future.”
These kinds of practical implications of data control are something Northland Māori tech company Te Hiku Media wants whānau to understand.
“What if AI was developed to fast-track and implement decisions in the Ministry of Justice?” asks CEO Peter-Lucas Jones (Te Aupōuri, Ngāi Takoto and Ngāti Kahu). “What’s the current data going to tell us? That at least 50 percent of people in jail must be Māori?”
His point is that, again, the data doesn’t tell us why Māori are in jail, and it doesn’t reveal the embedded bias which sees young Māori and Pākehā offenders treated differently for the same crimes.
He urges people to remember the taonga taken from Māori during the colonial era — from tools and art to heads and bones — and shipped off to museums and private collectors around the world. Many of those taonga are yet to be returned to Aotearoa. And that’s a reminder, he says, to ensure that this doesn’t happen to the taonga we still have, the greatest of which, in a data sense, are our reo and our whakapapa.
A huge amount of whakapapa and even Māori DNA data is currently being stored overseas in the databases of genealogy companies like Ancestry.com and 23AndMe. The issue, according to Te Hiku’s chief technology officer Keoni Mahelona (Kanaka Maoli), is that the data you’re handing over isn’t just yours — it’s collective cultural data.
Where whakapapa used to be held by our elders and passed down through generations of carefully protected kōrero and waiata, now one individual can make the choice to hand their whakapapa over to an American company with a few clicks of a button.
“Likewise,” says Keoni, “You’re doing the same when you’re, say, running your reo classes on Facebook. You’re essentially now giving Facebook your cultural knowledge and the data which is embedded in the language that you’re speaking.”
Peter-Lucas agrees with Te Taka that it’s imperative that we create our own digital real estate for our reo, because, simply put: “If our languages don’t have a place in the digital world, they won’t have a place in the future.”
But at the same time, he says: “We don’t want to enable our language to be mined for our cultural and scientific content, to create wealth for the very people that stole all the wealth and resources from our people in the first place.”
That’s why Te Hiku creates its own digital tools, like the bilingual transcription tool Papa Reo, a hugely successful standalone service not reliant on the big tech platforms.
“How many of our people don’t even have a place to live? And the same experience could be very real in the digital world unless we collectively take some leadership to ensure that our people have real estate to develop, innovate, and own technology, and not be the servants of someone else’s technology,” says Peter-Lucas.
“We must ensure that the next 25 years are not a re-versioning of the last 25 years, or the 25 years before that.”
The good news is that the digital world is a space Māori should be excited about, not afraid of, say these experts.
Keoni says we’re nowhere close to real artificial intelligence, and that AI is simply a term to describe machine learning. Therefore, he says: “These are just tools. Tools that we can use to serve the communities we represent.”
Te Taka agrees. “We just need young people confident enough to say: ‘I am the master of this, it is not the master of me. I want to build this in a Māori way.’”
We’re already seeing that kind of thinking in action.
Te Taka is helping build iwi solutions at home and influencing corporations abroad. Te Hiku is building sovereign spaces online to protect te reo. People like Chrissy are upskilling to pass knowledge on to their iwi and hapū. Many iwi and Māori data groups are also working to better manage and store their own data locally.
Keoni shares a Hollywood-worthy vision for the future: an inter-Indigenous data storage, computing, and governance network — “a League of Pacific Nations, where we all band together.”
He and many of his colleagues believe that could create a cloud storage solution capable of rivalling Amazon and Microsoft, which were recently chosen over local providers by Te Whatu Ora to host its new $4.5 million multi-cloud data project.
“New Zealand has a local cloud provider which is pretty good, but there’s no way it can compete because of scale,” says Keoni. “But by joining up we can help gain some of those economies of scale.”
One example, from Keoni’s close friend and colleague, Michael Running Wolf, is that if we required Facebook, Google Maps or Siri to pay for our language data, we’d get laughed out the door. But if every Indigenous tribe in the world did the same thing, it would be a different story.
“Now, all of a sudden, it’s a much more important market size and it makes more sense for them,” says Keoni. “Plus, you’d have millions of people saying ‘Yes, those languages need to be on these platforms, but you can’t just take their language and profit from it. You have to pay royalties to this co-op, which then go back to the Indigenous people.’”
The dangers of digital colonisation may not be obvious yet because, on the surface, technology seems to be advancing our reo and connections instead of suppressing them. But when you think about an American company, instead of our kaumātua, holding our whakapapa, it becomes clearer.
It becomes clearer still if you think about that company deciding how our reo should be spoken, or which of our stories are “correct”. Or, if robots really did take over the world, as many of us fear, with the data it has now, what would happen to Māori?
We have an opportunity to put our foot down early, because we’ve seen what’s happened in history, and what we need to do differently now. It’s not quite the Hollywood time travel version of a second chance. But it is a chance to learn from our history — and take control of our story and our future.
Siena Yates is an E-Tangata writer. This piece was made possible by the Public Interest Journalism Fund.
Thank you for reading E-Tangata. If you like our focus on Māori and Pasifika stories, interviews, and commentary, we need your help. Our content takes skill, long hours and hard work. But we're a small team and not-for-profit, so we need the support of our readers to keep going.
If you support our kaupapa and want to see us continue, please consider making a one-off donation or contributing $5 or $10 a month.