News River Express

News River Express is back with three articles worth your time, plus a bit of our own news.

We built this summer format so we can spend more time outside with our plants and our people, and still get the best links to your inbox.

This week is the topic everyone is talking about: AI, and where it meets Indigenous Data Sovereignty.

How Indigenous technologists are building data sovereignty into AI

BetaKit

Our CEO and Co-Founder Jeff Ward spoke at Mila's Indigenous AI Gathering in Montréal. The line that stayed with him came from Canada's Minister of AI and Digital Innovation, Evan Solomon:

"AI moves fast. Human hearts move slow."

The Minister came to an Indigenous-led space to present and to listen, weeks after Canada's national AI strategy launched. That's where trust starts, and it's how we build at Animikii: technology made with love starts with human hearts and lets love set the pace.

The BetaKit piece covers some of the conversations. Participants in the Indigenous Pathfinders in AI program pitched tools that keep data inside the community network. Diné geneticist Krystal Tsosie put the stakes plainly:

"[Data] determines what the system sees, what it optimizes for, who benefits, and who is harmed. Our community's data is already on this pathway, and the question is whether or not our community governs what it becomes."

Data centers are booming. Indigenous leaders want help protecting their lands.

Grist

AI infrastructure is hungry for land, energy, and water, and Indigenous delegates took that to the UN's Expert Mechanism on the Rights of Indigenous Peoples. A single hyperscale data centre can draw as much electricity as 100,000 households. Delegates called for free, prior, and informed consent, and some US tribes have passed outright moratoriums. As Sámi Parliament member Maren Storslett said:

"We must not only ask what AI can do, but what it should do."

Respecting First Nations knowledge on Wikimedia platforms

Wikimedia Australia

Wikimedia Australia and the Indigenous-owned firm Terri Janke and Company have released a guide for handling Indigenous knowledge across Wikipedia, Commons, and Wikidata. It matters for AI because Wikimedia trains many large language models, so errors and colonial framing scale with them. The guide uses a risk-assessment approach built on one question: who has the cultural authority to tell this story? As Terri Janke and Company's Shevaun Wright puts it:

"It's not about content moderation, but about content integrity."

And in Animikii news: we are testing ethical AI with an Indigenous community

We've partnered with CEIMIA on a first practical use case for ethical AI and data governance with an Indigenous community in Canada. It pairs two tools: our Niiwin platform, and the Data Roadmap CEIMIA built for the Global Partnership on AI.

Most talk about responsible AI stays at the level of principles. This pilot asks what those principles look like when a community runs them on its own data, with tools it controls. Niiwin carries OCAP®, CARE, and UNDRIP principles into how the system behaves. Testing CEIMIA's roadmap on Indigenous data should push the tool somewhere more inclusive, because community data raises questions government data never does.

It's a small, concrete pilot to help answer a bigger question: can these systems advance Indigenous rights, not just avoid harming them? That's what we're setting out to test. More information on that pilot here.

Want to talk it through? Book an AI conversation and Niiwin demo direct with our CEO Jeff: grab a time.