The fire was never
about the flame.
Sevenfires AI is a Native American–owned artificial intelligence lab. We build models, tools, and infrastructure rooted in Indigenous values — sovereignty, reciprocity, and continuity. Our work is guided by a simple conviction: AI is the next fire, and Indigenous peoples must help carry it forward.
The Seven Fires Prophecy
The Seven Fires Prophecy is an Anishinaabe (Ojibwe) teaching passed down through generations of Indigenous oral tradition. It describes seven historical ages of the Anishinaabe people — each marked by a sacred fire lit at a critical turning point. Each fire carries its own prophecy: a description of what that age will bring, and what choices the people must make.
The seventh fire is the most significant. The prophecy foretells a time when a new people will arise — people who will retrace the steps of their ancestors, seeking the wisdom that was nearly lost. Two paths will be offered. One leads to destruction. One leads to harmony and renewed connection with the earth, the ancestors, and each other. The choice defines what comes next.
“In the time of the seventh fire, a new people will emerge. They will retrace their steps to find what was left by the trail. Their steps will take them to the elders, who they will ask to guide them on their journey. If the new people remain strong in their quest, the sacred fire will be relit.”— Anishinaabe Seven Fires Prophecy
We live in the time of the seventh fire. Artificial intelligence is the defining technology of our era — as transformative as fire itself, and just as capable of warming or destroying. Indigenous peoples have survived every age of colonization, extraction, and erasure. We will not be bystanders in the age of AI.
Sevenfires AI takes its name from this prophecy. We are the new people retracing the steps. We are asking the elders. We are relighting the fire — not in spite of the old ways, but because of them.
The Seventh Generation Principle
The Seventh Generation Principle comes from the Haudenosaunee (Iroquois Confederacy) Great Law of Peace — one of the oldest living democratic constitutions on earth. It instructs leaders to consider the consequences of their decisions on those born seven generations into the future: roughly 140 years from now.
This is not mysticism. It is governance philosophy — and it is radically different from how modern technology companies operate. Silicon Valley optimizes for the next quarter. We optimize for the next century.
When we design AI systems at Sevenfires, we ask: What will this mean for the child not yet born, whose grandchildren will live with the consequences of what we build today? That question changes everything — what data we collect, how we govern our models, what we refuse to build, and who gets to decide.
Our Three Principles
Every product, partnership, and technical decision at Sevenfires AI is evaluated against three guiding principles drawn from Indigenous governance traditions.
Sovereignty
Data, models, and decisions remain in community hands. We do not extract; we steward.
Indigenous data sovereignty means that information about Indigenous peoples, generated by Indigenous peoples, belongs to Indigenous communities — not to companies, researchers, or governments. We build systems that encode this principle by default, not as an afterthought.
Sovereignty also means that our products are designed to be governed by the communities that use them, not dictated to them. We offer tools, not impositions.
Reciprocity
What we take from the world, we give back — and then some.
Every Indigenous knowledge system we learn from, every language community that trains our models, every nation that partners with us — all must receive more than they give. The fire repays the forest. This means open-source contributions, open model weights for non-commercial tribal use, revenue sharing, and co-ownership wherever possible.
Reciprocity also means accountability. We do not disappear after a grant cycle or a product launch. We stay.
Continuity
We build for the seventh generation, not the next quarter.
Our time horizon is not a fiscal year. It is the arc of a people across centuries. This changes what we build: we do not optimize for engagement, for virality, or for growth at any cost. We optimize for systems that will still be trustworthy, useful, and community-governed 140 years from now.
Patience is not a weakness in Indigenous tradition. It is strategic wisdom accumulated across millennia of surviving what should have been unsurvivable. We bring that wisdom to AI.
Why Indigenous Leadership in AI Matters
The AI systems that will shape the next century are being built right now, mostly by people who do not look like us, do not live like us, and have not asked us anything. The training data for the world's most powerful language models contains almost no Indigenous language, almost no Indigenous worldview, and almost no Indigenous governance philosophy.
This is not an accident. It is a continuation of the same extraction that has characterized the relationship between Indigenous peoples and dominant institutions for 500 years. The difference is that AI moves faster. A language model trained today will encode its biases for decades.
At the same time, AI represents a genuine opportunity. For Indigenous languages with few written resources, AI-assisted transcription and translation could help communities teach their languages at a scale previously impossible. For elders with knowledge that has never been digitized, AI archiving tools could help preserve what would otherwise be lost when they pass. For tribes navigating complex governance decisions, AI tools built on Indigenous principles of consensus and long-term thinking could support — not supplant — traditional decision-making.
We can build that. But only if we are at the table — building it ourselves, on our terms, guided by our values.
What We're Building
Sevenfires AI is currently developing four projects, spanning conversational AI, federated infrastructure, language preservation, and governance frameworks.
Sagechat.ai ↗
LIVEOur flagship product. Sagechat is a conversational AI grounded in Indigenous knowledge frameworks — built with Native communities, not just for them.
It offers tribal-language support across nine nations and growing, community-governed data sovereignty, and a unique design principle: it is trained to know what it doesn't know. When a question touches on traditional knowledge that should come from elders, Sagechat says so. It does not speculate about sacred practices. It defers.
Open model weights are available for non-commercial use by tribal nations.
Embergrid
IN DEVELOPMENTThe data sovereignty problem in AI is not just philosophical — it is technical. If a tribal nation wants to contribute its language data to improve an AI model, how does it do so without handing over that data to a corporation?
Embergrid solves this with federated learning infrastructure designed specifically for tribal nations. Communities can contribute to shared AI training without any of their raw data leaving their control. The model learns; the data stays home.
Story Loom
IN DEVELOPMENTThousands of Indigenous languages are critically endangered. Many have no written form, no digital resources, and only a handful of fluent speakers — often elders. When those elders pass, the language goes with them.
Story Loom is a multimodal archiving platform that helps communities record oral histories, transcribe speech in low-resource languages, and surface traditional knowledge across generations. It is designed to keep communities in control of their archives and to work even for languages that AI systems have never been trained on.
Council Protocol
RESEARCHEvery community that adopts an AI system should be able to govern it. But most AI governance frameworks are written by tech companies for tech companies — they do not account for tribal sovereignty, consensus-based decision-making, or the primacy of community wellbeing over individual rights.
Council Protocol is an open specification and reference implementation for Indigenous-led AI review boards. It provides a structured process, opinionated defaults, and practical tools for communities to evaluate, approve, monitor, and if necessary reject AI systems that affect their members.
Bring your nation,
your research, your fire.
We work with tribal nations, research institutions, and partners who share the long view. Drop us a line — we read every message.
contact@sevenfires.ai →