Blog

Stay up-to-date with what we are doing

Blog

Navigating Indigenous Data Sovereignty in the Age of AI

A discussion with members of our Indigenous Steering Committee

Posted on 1st of September 2026 by Stefaan Verhulst, Andrew Zahuranec, Jordan Harris, Maui Hudson, Chantal Kamgne, WarīNkwī Flores

Navigating Indigenous Data Sovereignty in the Age of AI
Navigating Indigenous Data Sovereignty in the Age of AI

By Stefaan Verhulst, Jordan Harris, Andrew J. Zahuranec, Maui Hudson, Chantal Kamgne, and WarīNkwī Flores

Indigenous communities face two related threats from AI: marginalization from digital advancement and extraction of their cultural knowledge and language without consent or benefit. One way to address these challenges is to create pathways for Indigenous communities to exercise meaningful control over how their data is collected, governed, and used.

Ahead of the International Day of the World’s Indigenous Peoples, The GovLab hosted a webinar featuring Indigenous experts Chantal Kamgne (Localizzz / EngageAfricaNLP), WarīNkwī Flores (Kinray Hub / Kara::Kichwa), and Maui Hudson (Te Kotahi Research Institute / Local Contexts) to explore how data commons can offer a middle path toward true data self-determination. The webinar was moderated by Stefaan Verhulst (The GovLab), and Guilherme Canela de Souza Godoi (UNESCO) provided closing remarks.

Three major themes emerged from that discussion, which we detail below: the ability of data commons to serve as a tool for self-determination; how there can be one-size-fits-all solution to governance because such arrangements become a power instrument of top-down pressure; and the need for large institutions to “walk the talk” when it comes to supporting Indigenous communities to take control of their data.

Beyond the Binary: Data Commons as a Tool for Self-Determination 

Steward Data Common Ecosystems—collaboratively governed ecosystems that pool responsible and ethical data access for the public good—offer a viable alternative to extractive “open” data practices. 

This idea was highlighted through several key exchanges about data ownership and community partnership design. Panelist WarīNkwī Flores emphasized that "data" itself is often a contested concept. For many Andean and Amazonian biocultural communities, it represents living ways of being, ways of doing, ways of knowing, ways of not knowing, and embodied biocultural practices rather than numerical variables and mere FPIC checkboxes.

Without intentional governance, AI risks repeating historical patterns of apartheid engagement in which communities are invited as cultural symbols rather than equity shareholder negotiators. Because AI only accesses what is already digitized, bridging the digital divide requires operationalizing community autonomy’s rule of law—including the choice of voluntarily isolated populations who may not want digital transformation intervention at all.

This also means resisting the tendency to make technology solutionism the starting point. As Maui Hudson emphasized, the goal should not be to assemble Indigenous language resources for AI, but to strengthen those resources for Indigenous communities themselves. AI may be a useful tool toward that end, but it should follow community priorities rather than define them.

 

Toolboxes Over One-Size-Fits-All Governance 

There is no single legal framework that fits every territory. Throughout the discussion, panelists advocated for a "toolbox approach," combining global instruments such as UNDRIP, IEEE std 2890-2025, and renewing the attention of CARE Principles (Collective Benefit, Authority to Control, Responsibility, Ethics) to redirect data actors to territorial data governance with practical tools such as Traditional Knowledge (TK) Labels developed by Local Contexts, Kara::Kichwa Data Sovereignty, Sovereign Data Supply Chain, and Place-based Relational Infrastructure - Automatization of FPIC by Kinray Hub and Partners.

Key practical takeaways included:

  • Platform Fine Print: Standard commercial platforms frequently claim ownership over uploaded data in their terms of service, requiring communities to carefully select hosting environments.

  • BioKultural Impact Assessment (BKIA): Frameworks like those created by Kinray Hub allow communities to assess their own readiness, priorities, and governance rules before formal Free, Prior, and Informed Consent (FPIC) processes.

  • Community-Defined Priorities: As Chantal Kamgne noted using Cameroonian language examples, projects must solve immediate local needs—such as accessing emergency services or restoring traditional land management—rather than satisfying external tech metrics.

 

Walking the Talk: The New Commons Incubator 

Closing remarks from UNESCO’s Guilherme Canela highlighted how AI has forced the convergence of multilingualism, open solutions, and data governance. Aligned with the International Decade of Indigenous Languages, the New Commons Incubator aims to translate these theoretical rights into concrete proposals by supporting Indigenous-led teams in building fundable, community-governed data platforms.

***

Data sovereignty is ultimately a practice, not just a legal status. Exercising it requires communities to have the capacity, institutions, tools, and agency to determine how, and whether, their data is used. By building local technical capacity and supporting translational leaders, the New Commons Incubator aims to help turn formal rights into practical authority over communities’ digital futures.

If this work interests or inspires you, please feel free to contact us at [email protected] and let us know.

 

Back to the Blog

Supported by