Indigenous Data Sovereignty — CARE Principles, IEEE 2890-2025, and operational anchors in Peru and Canada

Global (with operational anchors in Peru, Canada, Mexico, Aotearoa New Zealand)

Content

Indigenous Data Sovereignty (IDSov) is the right of Indigenous Peoples to own, control, access and possess data that derive from them, and which pertain to Nation membership, knowledge systems, customs, or territories. The framework is articulated through the CARE Principles, operationalised in IEEE 2890-2025, and anchored in the UN Declaration on the Rights of Indigenous Peoples (UNDRIP).

This unit captures the framework and the operational anchors — Indigenous-led deployments that have moved from framework articulation to substantive community outcomes.

The CARE Principles (Carroll et al. 2020, Gaborone)

Drafted: November 2018, International Data Week and Research Data Alliance Plenary, Gaborone, Botswana Co-leads: Stephanie Russo Carroll (University of Arizona) and Maui Hudson (University of Waikato, Aotearoa New Zealand) Citation count: 1,808 (Carroll et al. 2020, Data Science Journal) — substantive academic uptake Translations: Spanish, Vietnamese, Māori, German, Khmer

The four principles:

• Collective benefit — data ecosystems should enable Indigenous communities to derive benefit from their data
• Authority to control — Indigenous Peoples’ rights and interests in their data are recognised
• Responsibility — those working with Indigenous data are accountable to Indigenous communities
• Ethics — Indigenous worldviews and rights are foregrounded in data work

The “FAIR + CARE” framing. CARE Principles were explicitly developed to complement the FAIR Principles (Findable, Accessible, Interoperable, Re-usable). The GIDA framing: “Existing principles within the open data movement (e.g. FAIR) primarily focus on characteristics of data that will facilitate increased data sharing among entities while ignoring power differentials and historical contexts.”

IEEE 2890-2025 — Provenance of Indigenous Peoples’ Data

PAR Approval: June 3, 2020. Board Approval: September 10, 2025. Published: November 14, 2025 (5-year development process). Status: World’s first global standard for the provenance of Indigenous data Format: IEEE Recommended Practice (industry specification, not binding regulation)

Working Group Chair: Stephanie Russo Carroll (University of Arizona). IEEE Program Manager: Malia Zaman. Standard Committee: SSIT/SC — Social Implications of Technology Standards Committee. Working Group: ID-WG — Indigenous Data Working Group.

What it does: Operationalises the CARE Principles into industry specifications by detailing the process for describing and recording the provenance (origin and history of ownership) of data derived from Indigenous Peoples, lands, and waters. Critically: explicitly applies to AI/ML industries (per Udall Center event description, February 2026): “Within data-generated industries, such as artificial intelligence and machine learning, as well as biodiversity research and innovations in genomic science, this standard enhances the integrity of that data.”

Recognition: GIDA recognised IEEE 2890-2025 as including the active participation of Indigenous members in the development process, with Indigenous Caucus, consultation with Indigenous Peoples, and ensuring open access. Webinar (February 18, 2026) led by Stephanie Carroll, Jane Anderson, Randy Akee, Camille Callison, Maui Hudson, Max Liboiron — hosted by Indigenous Data Exchange, Collaboratory for Indigenous Data Governance, IndigeLab Network, CBIKS and ENRICH.

Operational anchors — Indigenous-led deployments

PolArctic Sanikiluaq Nunavut (Two-Eyed Seeing operational pattern)

Approach: Two-Eyed Seeing (Etuaptmumk) — Albert Marshall’s Mi’kmaw concept that integrates Indigenous Knowledge and Western science as equals. First AI to operationalise this for a Canadian Inuit community. Inuktitut + English output.

Application: Fisheries management + climate adaptation + shoreline protection monitoring in Sanikiluaq, Nunavut (Qikiqtaaluk Region, Belcher Islands).

Peer-reviewed: Iravani, R., Biagi, M., Laforest, S., Lee, K., Isaacman, L., Chen, Z., An, C. (2025). Protecting shorelines in Canadian Indigenous communities. Marine Pollution Bulletin 219:118310. doi:10.1016/j.marpolbul.2025.118310.

See units/polarctic-sanikiluaq-mariculture-ai.md for full unit.

Salmon Vision BC (Gitanyow Fisheries Authority + SFU)

Approach: AI-integrated fish weir monitoring. Combining AI with centuries-old Gitanyow fishing weir technology. Indigenous-led methodology with cutting-edge computer vision.

Performance: ~90% coho accuracy / ~80% sockeye accuracy. 12 species. 500,000+ frames annotated.

Peer-reviewed: Frontiers in Marine Science 2023.

Operational locations: Bear River + Kitwanga River fish counting weirs, BC Central Coast.

See units/salmon-vision-bc-wild-salmon-ai.md for full unit.

Indigenous Navigator Peru (ONAMIAP-IWGIA, 2023-2025)

Lead organisations: ONAMIAP (Organización Nacional de Mujeres Indígenas Andinas y Amazónicas del Perú — National Organisation of Andean and Amazonian Indigenous Women of Peru) + IWGIA (International Work Group for Indigenous Affairs) Funding: European Union Communities: Quechua and Asháninka in Peru Period: 2023-2025 Published: December 2025 (case study), updated April 2026

Concrete outcomes:

• Documented collective rights challenges using Indigenous Navigator framework
• Identified priorities linked to water scarcity, territorial integrity, food sovereignty, and local governance
• Moved from data collection → small-grant projects → municipal ordinance advocacy
• Specific community initiatives supported: water conservation, protection of natural springs, reforestation, fish farming, seed recovery, territorial mapping

Substantive significance: This is a concrete Indigenous-led deployment with policy outcomes (municipal ordinances). Not just framework articulation.

FNIGC — First Nations Information Governance Centre (Canada)

Founded: 1990s Status: Established national-scale Indigenous data governance in Canada Annual Report 2024-2025 referenced as the operational anchor for First Nations data sovereignty OCAP® principles (Ownership, Control, Access, Possession) — the Canadian Indigenous data governance framework. Established 1998 at the National Steering Committee meeting of the First Nations and Inuit Regional Longitudinal Health Survey; the acronym began as “OCA” (attributed to NSC member Cathryn George) and a “P” was added by the committee. FNIGC incorporated 22 April 2010. OCAP® and PCAP® are registered trademarks with the Canadian Intellectual Property Office (CIPO); logos granted August 2014, acronyms August 2015. PCAP® French-language version launched 2024.

National Inuit Strategy on Research (NISR, ITK 2018) — the substantive Inuit data-sovereignty framework, analogous to OCAP for First Nations. Five priority areas including (iv) ensuring Inuit access, ownership, and control over data and information. NISR is the operative citation in SIKU / Arctic Eider Society + SmartICE + Eweal primary-source literature (per Nowosad et al. 2025 Arctic Science). https://www.itk.ca/wp-content/uploads/2018/04/ITK_NISR-Report_English_low_res.pdf

First Nations Data Governance Strategy (FNDGS): Submitted to Indigenous Services Canada in 2025 (AFN Resolution 30-2025). Community-driven, Nation-based national strategy. Plans regional information governance centres (one per region). See units/fnigc-first-nations-data-governance-strategy.md for full unit.

NCIAF — National Circle for Indigenous Agriculture and Food (Canadian institutional vehicle)

Launched: 13 June 2024 with FCC support. The only named Indigenous-led agrifood organisation with FCC backing and national scope. MOU with CFA signed 2025. President and CEO Kallie Wood; Senior Director Gabrielle Fallu.

See units/nciaf-indigenous-agrifood-canada.md for full unit.

Te Mana Raraunga — Māori Data Sovereignty Network (Aotearoa NZ)

Positioning: Māori Data Sovereignty / Māori Data Governance “Tino Rangatiratanga Raraunga” — Māori data sovereignty as self-determination “State of the Nation” Report 2025 (Taiuru & Associates, September 2025) — annual state-of-Māori-data-sovereignty assessment Context: Aotearoa NZ has substantive Māori-led AI/data infrastructure development, including Te Reo Māori language AI, Māori data governance legislation

UNESCO Indigenous AI Report (2023, updated 2025)

“Indigenous People-Centered Artificial Intelligence: Perspectives from Latin America and the Caribbean” Region covered: 800+ Indigenous Peoples in LAC; 10%+ of world’s Indigenous population; ~30% in extreme poverty; 40% with basic computer skills

Five Mexico case studies:

1. UAM researchers — text-strings and AI proposal to catalogue and identify form, aesthetic, and iconographic elements of indigenous garments of the Altos de Chiapas region
2. Technological Institute of Coatzacoalcos (Veracruz) — NLP bot to assess pronunciation of indigenous languages
3. Technological Institute of Oaxaca — interactive learning app for Tu’un Savi (Mixtec) language using visual computing
4. UNAM (IIMAS) — automatic translation between 11 indigenous language families using ML/deep learning (Wixarika, Náhuatl, Yorem nokki, P’urhecha, Mexicanero, Spanish)
5. f<A+i>r + CEPIADET + Oaxaca Human Rights Defender — conversational agent for indigenous language interpreters; the proposed system was inadequate according to the evaluation — a substantive honesty about implementation challenges

The fifth case is particularly worth knowing: it failed, and that failure was reported honestly. Worth surfacing in any talk about IDSov operationalisation.

AGUAPAN — Association of Guardians of the Native Potato of Central Peru

Founded: July 2015 Type: Custodian farmer network — farmer-led biodiversity conservation Activities: Protect varietal richness of Central Peruvian Andean potatoes; exchange of landraces across Peru’s many agro-ecological zones; 5th Annual Assembly convened Funding: McKnight Foundation / Collaborative Crop Research Program (CCRP) Related initiative: Potato Park (Pisac, Peru) — climate-resilient potato breeding

Worth distinguishing: AGUAPAN is not primarily a data sovereignty project — it’s a custodian-farmer biodiversity network. But it intersects with IDSov because the genetic resources and traditional knowledge at stake are Indigenous data in the broader sense (genetic resources, traditional cultivation knowledge). The WIPO Treaty on Intellectual Property, Genetic Resources and Associated Traditional Knowledge (2024) explicitly addresses this intersection.

Related international mechanisms

UNDRIP — UN Declaration on the Rights of Indigenous Peoples: Article 3 (self-determination), Article 31 (traditional knowledge). The legal anchor for IDSov.

EMRIP 18th Session (July 2025): Expert Mechanism on the Rights of Indigenous Peoples adopted draft study “Right of Indigenous Peoples to Data, Including with Regard to Data Collection and Disaggregation,” framing data as “cultural, strategic and economic resource.” Issued Advice No. 18, calling on states to recognise and protect Indigenous data sovereignty through laws, policies, and frameworks.

WIPO Treaty on Intellectual Property, Genetic Resources and Associated Traditional Knowledge (2024): First ratifications by Malawi (December 2024) and Uganda (July 2025). Will enter into force after 15 ratifications. Distinct from CARE / IEEE 2890 but addresses the genetic resource / traditional knowledge intersection.

Cali Fund (CBD): Fair and Equitable Sharing of Benefits from use of Digital Sequence Information on Genetic Resources. Growing global recognition.

GIDA “CARE Directs Us Home” communiqué (August 2025): Warns against treating CARE as static compliance checklist. Reasserts CARE as high-level guidance directing researchers toward community-specific protocols.

Global Indigenous Data Sovereignty Conference (Canberra, April 2025): 250+ Indigenous representatives; three foundational actions: advancing Indigenous data priorities, addressing systemic barriers, building Indigenous data capability.

Critical context

The framework is more developed than the Indigenous-led AI deployment. CARE Principles (2018/2020), IEEE 2890-2025, EMRIP 18, GIDA communiqués — the framework layer is substantive. The Indigenous-led AI deployment layer is thin in agrifood specifically; the UNESCO report’s five Mexico case studies are the most concrete operational anchors in agriculture-adjacent AI.

The geographic distribution matters. Aotearoa NZ (Te Mana Raraunga, Māori data sovereignty), Canada (FNIGC, OCAP®), Peru (ONAMIAP-IWGIA Indigenous Navigator), Australia (FNIGC’s analog), Mexico (UNESCO report), Asia (AIPP), Africa (African Commission on Human and Peoples’ Rights, 9 August 2025 statement) — IDSov is global but uneven in operational scale.

The “be FAIR and CARE” framing is the key insight. CARE complements FAIR rather than replacing it. Talks about Indigenous data governance should engage with both — CARE addresses people and purpose; FAIR addresses technical interoperability. Together they form a complete framework.

Worth naming honestly: Indigenous-led AI deployment in agrifood specifically is under-deployed relative to vendor-led or state-led AI deployment, but substantive operational deployments do exist: PolArctic Sanikiluaq (peer-reviewed 2025), Salmon Vision BC (peer-reviewed 2023), NEDA AI tool (CanNor REGI-AI), Inuvik Tech Society (CanNor IDEANorth), Indigenous Pathfinders in AI (Mila 2025 cohort). The gap is funding-instrument-specific (no Indigenous-led-agrifood-AI-specific federal programme), not framework-substantive or operational-deployment-substantive. The framework-deployment funding-instrument gap is G-313.

The Canadian agrifood analytical voice — Ruder + Wittman + collaborators — is substantively important:

• Ruder, S.-L. & Wittman, H. (2025). Agricultural data governance from the ground up: Exploring data justice with agri-food movements. Big Data & Society. Cited 22 times. 40 workshops + 26 conferences across Western Canada, US, Latin America (2020-2023).
• Ruder, S.-L. (2025). Theorizing agricultural data policy and governance. Journal of Peasant Studies. Cited 14 times.
• Duncan, E. et al. (incl. Ruder). (2026). The political economy of agricultural data governance. Journal of Rural Studies. Cited 6 times.
• BCACARN Ethical Data Governance Toolkit (2024) — first documented Canadian agrifood ethical data governance toolkit.

Ruder-Wittman are the substantive Canadian agrifood data-governance analytical voice — distinct from Neethirajan (Dalhousie; AI-deployed-on-the-farm voice), Dara (U Guelph; greenhouse AI risk), Charlebois (Dalhousie; pragmatic-critical).

Mila’s substantive Indigenous AI work is institutional infrastructure, not a CSA joint-committee standard (correction to the constraint/critical scout §5 framing):

• Indigenous AI Advisory Circle (Mila + CIFAR; 6 Indigenous leaders; convening function)
• Indigenous Pathfinders in AI (Mila summer program; 2nd cohort 2025: 21 participants / 8 AI prototypes; capacity-building function)
• Pathways to AI: An Introduction for Indigenous (Mila + FNTC course; education function)
• Mila Indigenous AI Gathering 2026 (July 15-16 2026, Montreal) — operating institutional anchor. Dana Tizya-Tramm (Vuntut Gwitchin Former Chief; Executive Director, Nadlii Society) named speaker.
• Mila AI Policy Fellowship 2026-2027 — Indigenous AI thematic area.

CSA Group’s relevant artefact is internal staff Indigenous Engagement Guidelines (not a published standard). No published Mila-CSA Group standard on Indigenous data sovereignty surfaced as of July 2026 — reported as honest absence.

See units/mila-indigenous-ai-infrastructure.md for full unit.

What this unit is doing in the taxonomy

First unit in the field guide with data-rights-framework: Indigenous-sovereign as primary value. Fills the v4 schema gap identified in the data cooperatives / commons deepening.

Anchors the data-substrate × Indigenous-sovereign cell with the CARE Principles framework, IEEE 2890-2025 standard, and concrete operational anchors in Peru, Canada, Mexico, and Aotearoa NZ.

Distinct from:

• IPES-Food corporate concentration lens (farm-data-ownership-critical.md) — different critical voice; IDSov is rights-based, IPES-Food is corporate-concentration.
• JoinData (joindata-netherlands.md) — Dutch cooperative cultural fit, not Indigenous-led.
• OADA (oada-open-ag-data-alliance.md) — open-source standards, not Indigenous-led.
• CGIAR FAIR Principles (open-data-ecosystem.md) — multilateral open data, complementary to CARE.
• Cornell Atkinson IDSov cluster (cornell-atkinson-idsov-cluster.md) — US land-grant academic IDSov work.
• Canadian IDSov + AI in Agrifood (canadian-indigenous-data-sovereignty-agrifood.md, new) — consolidates the framework + deployment + institutional + analytical-voice layers for Canadian context.

Why it matters for talks

• The CARE Principles + IEEE 2890-2025 are the substantive framework anchors.
• The Indigenous Navigator Peru case is the most concrete operational example in agrifood-adjacent context.
• The UNESCO five Mexico case studies are concrete Indigenous-led AI deployment.
• The “FAIR + CARE” framing is the substantive synthesis — talks should engage with both.
• The “be FAIR and CARE” multilingual translations matter for global agrifood AI deployment.

Critical context

• Framework articulation > deployed Indigenous-led AI infrastructure in agrifood specifically. Honest about the gap.
• Indigenous Navigator Peru and FNIGC are the most concrete operational anchors.
• UNESCO Mexico case studies are concrete Indigenous-led AI deployment; the f<A+i>r / CEPIADET case failed and that failure was reported.
• The 1,808 citation count on Carroll et al. 2020 is substantive academic uptake.
• GIDA’s “CARE Directs Us Home” (August 2025) is the most recent substantive statement on CARE implementation.