Internet Explorer 11 is not supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

The Protection That Tribal Nations’ Data Needs

AI poses a threat to Native sovereignty over cultural knowledge. Tribal nations should have the authority to govern how their data is collected, stored, interpreted, shared and used.

Stephanie Russo Carroll, an associate professor of public health at the University of Arizona
Stephanie Russo Carroll, an associate professor of public health at the University of Arizona, co-founded the U.S. Indigenous Data Sovereignty Network and directs the Collaboratory for Indigenous Data Governance. (Kris Hanning/University of Arizona/TNS)
Public agencies are moving quickly to adopt artificial intelligence. Federal guidance now encourages agencies to use and procure AI tools, while state and local governments are increasingly experimenting with systems that are already shaping public services, records, research and decision-making.

These efforts may have the intention of improving governance. But when it comes to cultural knowledge and data, AI technologies pose important questions about tribal nations’ sovereignty and governing authority that are too often overlooked by other government actors.

For tribal nations, these are old questions: whether outside institutions will once again decide how Native knowledge is used or if they will recognize tribal nations’ authority. Too often Native peoples’ records, policies, archives and public narratives were collected to serve colonial authority rather than Native self-determination.

AI poses a similar threat. It can repurpose records collected without consent. It can absorb cultural materials into systems built for outside purposes. It can produce outputs that sound authoritative while distorting the realities of distinct tribal nations.

That does not mean Native peoples are approaching AI only with fear. At the recent U.S. Indigenous Data Sovereignty and Governance Summit in Tucson, I heard Native leaders, scholars, technologists and community advocates asking not simply whether AI should be embraced or rejected but how tribal nations can govern it.

That begins with a simple premise: Native knowledge is not just data. It spans government records, language materials, photographs, songs, oral histories and more. In many cases, researchers collected these materials under conditions that would not meet any serious standard of informed consent today.

Now, AI systems scrape this information across many communities. But the Native data that these systems collect raises a distinct sovereignty issue because it is tied to Indigenous self-determination. Native materials are often treated as available for reuse by anyone, but institutional access is not the same as tribal consent. Converting these materials into digital files, searchable collections or training data does not authorize their use in AI systems.

This is why policymakers and public agencies must treat Indigenous data sovereignty as a governing principle in AI funding, procurement, research approval and deployment, not a diversity add-on. Indigenous data sovereignty means that tribal nations have the authority to govern how their data is collected, stored, interpreted, shared and used.

No one expects federal, state or local governments to give outside institutions unrestricted authority over their records, infrastructure or public responsibilities. Tribal nations are also governments, and it should be required to treat them that way.

Research agreements can protect sovereignty, or they can quietly transfer power away from Native communities. Vendors may not understand tribal governance. Agencies may consult tribal nations only after a system has already been designed, a contractor selected or a data set acquired. At that point, consultation becomes notice. Notice is not governance.

There is a better way forward. Existing Indigenous data frameworks offer guidance about whether data practices produce collective benefit, respect Indigenous authority and follow ethical obligations. Other frameworks emphasize ownership, control and access over Indigenous data. Together, these approaches shift the question from whether data can be used to who has the authority to decide how Indigenous data is collected, interpreted, shared, stored and reused.

Policymakers and public agencies should apply such standards before AI systems are trained or deployed. Any AI system that uses Native data or significantly affects tribal nations should require an Indigenous data sovereignty plan developed through meaningful tribal consultation. That plan should explain who can access the data, where it will be stored, how long it will be kept and whether vendors are allowed to use it. Such a plan should also spell out the consequences if the terms are violated.

There is real opportunity here. AI could help tribal nations organize dispersed archival records. It could support language learners through tools designed under tribal authority. It could help improve service delivery or analyze infrastructure needs. But those uses are beneficial only if tribes determine the purpose, control the data, set the limit and determine what AI is allowed to become.

Native peoples have been at this crossroads before. They know what extraction looks like. They also know adaptation, governance and survival. The question is not whether AI will shape the future. The question is whether the authority tribal nations already hold will be respected to shape AI as well.

Kerri J. Malloy is an assistant professor of Native American and Indigenous studies at San José State University whose research focuses on Indigenous memory, genocide studies, public institutions and tribal sovereignty.



Governing's opinion columns reflect the views of their authors and not necessarily those of Governing's editors or management.