CARE principles

The open data movement often starts with a useful technical question: can data be found, accessed, combined, and reused? The FAIR principles answer that question. FAIR stands for findable, accessible, interoperable, and reusable. It helps data move.

The CARE principles ask a different set of questions. Should the data move? Who benefits if it does? Who has authority over the data? Who is responsible for what happens after reuse? Whose ethics define harm and future use?

The Research Data Alliance International Indigenous Data Sovereignty Interest Group published the CARE Principles for Indigenous Data Governance in 2019 through the Global Indigenous Data Alliance. The paper argues that open data and open science often focus on the properties of data while leaving power differences and historical context outside the frame. For Indigenous data, that omission is not a small detail. It can decide whether data supports self-determination or becomes another form of extraction.

CARE stands for Collective Benefit, Authority to Control, Responsibility, and Ethics.

Collective Benefit asks whether data ecosystems are designed so Indigenous Peoples derive benefit from the data. That benefit can mean better governance, local innovation, stronger services, or value created in ways grounded in Indigenous worldviews.

Authority to Control asks whether Indigenous Peoples' rights and interests in Indigenous data are recognised. It includes rights to free, prior, and informed consent, rights to data for governance, and rights to develop cultural protocols for how data are collected, represented, accessed, and reused.

Responsibility asks whether people using Indigenous data can show how that use supports self-determination and collective benefit. It is about relationships, evidence, capacity building, data literacy, infrastructure, and respect for Indigenous languages and worldviews.

Ethics asks whether Indigenous Peoples' rights and wellbeing are the primary concern across the data life cycle. Ethical data do not portray Indigenous Peoples, cultures, or knowledges through deficit. Ethical processes address power imbalances and include representation from the communities to whom the data relate.

For AI, CARE is especially concrete. A model can turn a dataset into a classification, a translation, a recommendation, a risk score, or a training example. CARE says governance must travel with that movement. A dataset that is technically reusable may still be wrong to reuse if collective benefit, authority, responsibility, and ethics are missing.

Source: Research Data Alliance International Indigenous Data Sovereignty Interest Group. September 2019. "CARE Principles for Indigenous Data Governance." The Global Indigenous Data Alliance. GIDA-global.org.