Indigenous Nations Are Shaping the Future of AI on Their Own Terms

From safeguarding endangered languages to strengthening healthcare and making tribal law more accessible, Native American nations are exploring artificial intelligence while asserting control over their data, knowledge and digital futures.

periences of integrating artificial intelligence into governance, healthcare, language preservation and public services.

While the potential of AI was widely acknowledged, participants also emphasised the need for strong safeguards around privacy, cybersecurity, cultural knowledge and institutional oversight.

For Indigenous governments, the challenge is not simply whether to adopt artificial intelligence, but how to ensure that its development and deployment remain consistent with community priorities.

Preserving a Language Before Its Living Voices Disappear

For the Choctaw Nation of Oklahoma, artificial intelligence is becoming part of an effort to protect a language facing a profound generational challenge.

Despite being the third-largest federally recognised Native American tribe in the United States, the Choctaw Nation has approximately 300 first-language speakers remaining, according to Zachary Harbort, the Nation’s IT manager for low-code platforms.

The urgency is personal as well as institutional. Harbort shared that the last first-language speaker in his own family had passed away during his lifetime.

Working with his team, Harbort has been digitising interviews with elders, historical recordings and other primary materials to create a structured repository of Choctaw linguistic knowledge.

He describes the resulting archive as a digital seed vault — a resource intended to preserve linguistic material and support future generations of learners.

Beyond storing recordings and documents, the initiative is exploring how AI-powered applications can make the language more accessible in everyday life.

One prototype developed using AI-assisted coding allows users to point a device at an object and identify its name in Choctaw, provided that the corresponding word exists in the dictionary used by the application.

A cup, a vehicle or another familiar object can thus become an opportunity for language learning.

The application demonstrates how emerging technologies might help connect digitised linguistic resources with interactive educational experiences.

Yet the initiative also reveals an important tension. Harbort acknowledged concerns about AI technologies, particularly systems that introduce listening and data-collection capabilities without adequate consideration of their consequences.

The Choctaw approach reflects a distinction between using technology to extract linguistic knowledge and developing tools that help a community sustain its own language.

Digitisation alone cannot guarantee a language’s survival. Continued transmission between generations, community participation and living cultural practice remain essential. However, carefully designed digital tools may provide valuable support to these efforts.

Returning Time to Healthcare Workers

Within Cherokee Nation Health Services, another application of artificial intelligence is addressing a different concern: the administrative demands placed on healthcare professionals.

For clinicians, patient care extends beyond consultations. Medical documentation, case histories, treatment notes and administrative records can consume considerable time throughout the working day.

Darylin Doublehead, a former clinician who joined the Cherokee Nation health IT team in 2024, has been involved in piloting Oracle’s Clinical AI Agent to reduce this burden.

The system uses ambient listening technology to assist with clinical documentation. It processes conversations between patients and healthcare providers and generates draft notes containing relevant information, including symptoms, medical history, clinical observations and treatment plans.

Healthcare professionals remain responsible for reviewing and correcting the generated documentation.

At TribalNet 2025, Doublehead emphasised that the technology was intended to support clinicians rather than replace their professional judgement.

According to figures shared during the conference, the system had reached 92 users and processed at least 4,864 recordings across the Cherokee health system.

The reported average time saving was approximately two minutes per patient consultation. For a clinician seeing 20 patients in a day, that could amount to around 40 minutes.

Although the saving may appear modest when measured per consultation, its cumulative impact can be meaningful for professionals managing demanding schedules.

Doublehead connected these efficiencies to the everyday lives of healthcare workers, including the possibility of spending more time with their families.

The case illustrates an approach to AI adoption in which institutional efficiency is evaluated alongside human wellbeing.

It also underscores the importance of safeguards. Healthcare recordings contain sensitive personal information, making consent, confidentiality, secure processing and professional oversight essential considerations in any such deployment.

Making Cherokee Law More Accessible Through AI

Artificial intelligence is also being explored as a means of navigating the complexity of Indigenous legal systems.

The Cherokee Nation maintains an extensive body of law encompassing tribal codes, legislative amendments, executive orders, court decisions and treaties.

According to Cherokee Nation Chief Information Officer Paula Starr, the Nation’s tribal code alone exceeds 1,700 pages, without accounting for all subsequent amendments.

Much of this legal information exists across different platforms, documents and formats, making it difficult to locate and interpret the most relevant provisions.

The challenge affects not only Cherokee citizens seeking to understand their legal rights and obligations, but also legal professionals working within the Nation’s institutions.

To address this, Cherokee Nation has developed an AI-powered legal assistant using Microsoft Copilot.

The system is designed to search across multiple sources of legal information and help users navigate the Nation’s legal framework more efficiently.

Its development required detailed work to organise and understand the underlying documents. Starr explained that interns from the Massachusetts Institute of Technology assisted with the project, developing substantial familiarity with the legal materials.

The proposed system is intended to support citizens, lawyers and staff within the Office of the Cherokee Nation Attorney General once it receives approval for deployment.

Users would also be able to provide feedback to improve its responses.

The application remains significant beyond its technical functions.For Indigenous nations, legal systems are central to governance and self-determination. Making tribal law more accessible can help strengthen the relationship between citizens and their governing institutions.

However, AI-generated legal information must be treated carefully. Laws can change, interpretations may be contested, and inaccurate outputs could have serious consequences. Such tools therefore require authoritative sources, regular updates and human legal oversight.

The Question of Indigenous Data Sovereignty

Across these three initiatives, a common concern emerges: the relationship between technological innovation and Indigenous control over information.

Language archives may contain the voices and knowledge of elders. Healthcare platforms process confidential patient records. Legal systems hold information fundamental to tribal governance.

In each case, the value of AI depends not only on what the technology can do, but on how the underlying information is collected, stored, processed and governed.

At TribalNet 2025, Cherokee Nation CIO Paula Starr cautioned that artificial intelligence introduces risks of data exposure that demand a different approach from conventional technology deployments.Her concerns reflect a wider debate around Indigenous data sovereignty — the principle that Indigenous peoples should have authority over data concerning their communities, territories, institutions and knowledge systems.

For Indigenous governments, reliance on external technology providers can raise difficult questions about ownership, consent, access and long-term institutional control.

These issues become particularly important when digital systems involve cultural knowledge or community information that may not have been intended for unrestricted use.The initiatives presented at TribalNet suggest that technological adoption and sovereignty need not be opposing objectives, provided that Indigenous institutions retain meaningful control over the systems they choose to implement.