Between 20 April and 1 May 2026, the 25th Session of the United Nations Permanent Forum on Indigenous Issues (hereinafter, the Permanent Forum) took place.
The Permanent Forum is an advisory body to the United Nations Economic and Social Council (ECOSOC), established in the year 2000 with the mandate to examine issues pertaining to Indigenous Peoples in the areas of economic and social development, culture, the environment, education, health, and human rights more broadly.
It is composed of 16 members: eight proposed by governments and elected by the Council; eight appointed by the President of the Council, following official consultation with regional groups through their coordinators, based on extensive consultations with Indigenous Peoples’ organizations and taking into account the world’s diversity and geographical distribution.
At each annual session, the Permanent Forum examines progress and challenges in the implementation of the United Nations Declaration on the Rights of Indigenous Peoples.
During the most recent session, several documents were presented and discussed, including a paper authored by Hindou Oumarou Ibrahim entitled “Indigenous Peoples and Artificial Intelligence.”
Background
Emerging digital technologies, and artificial intelligence (AI) in particular, generate positive expectations but also raise concerns regarding their potential to deepen existing problems, notably current economic and political inequalities.
For Indigenous Peoples worldwide, the situation is highly alarming, as algorithmic biases open the possibility of eventual digital ethnocide—that is, their disappearance or extreme reduction within the databases that underpin AI.
In September 2024, the United Nations General Assembly approved the so-called Pact for the Future, which includes two annexes: the Global Digital Compact and the Declaration on Future Generations.
The Pact aims to achieve five objectives: closing digital divides; expanding inclusion in the digital economy; fostering an inclusive, open, and secure digital space; promoting responsible and equitable data management; and enhancing international governance of artificial intelligence so that it benefits humanity as a whole.
Explicitly, it assumes the recognition, respect, promotion, and protection of the rights of Indigenous Peoples, their territories, lands, and ecosystems, as well as their traditions, spiritual beliefs, and ancestral knowledge, guaranteeing their right to participate in decision-making on matters affecting their rights, in accordance with international human rights law.
Its text incorporates specific paragraphs concerning Indigenous Peoples and people of African descent, including the commitment to “Foster synergies between science and technology and traditional, local, Afro-descendant, and Indigenous knowledge, systems, practices, and capacities.”
Furthermore, the United Nations Educational, Scientific and Cultural Organization (UNESCO) adopted in 2021 the Recommendation on the Ethics of Artificial Intelligence, the first global framework on AI ethics, which includes principles such as human dignity, social justice, cultural diversity, inclusion, transparency, and the protection of human rights.
This instrument contains several points of particular relevance to Indigenous Peoples:
- The need to protect vulnerable groups against algorithmic biases;
- The importance of respecting cultural identities and local knowledge;
- The obligation to ensure the inclusion of marginalized communities in AI governance;
- The explicit recognition of the risk that AI may amplify structural discriminations.
Risks and Possibilities
In the context of the growing expansion of this issue, whose relevance is evident, the Permanent Forum has received and analyzed the document under discussion, which was presented as a significant input to facilitate the analysis of the current situation, accompanied by recommendations directed at various public and private actors.
One of the principal merits of the text lies in its compilation of concrete international and national examples that States—and especially Indigenous organizations—are promoting to harness new technology in the defense and advancement of their rights, while mitigating or, where appropriate, eliminating its adverse effects.
The document begins by highlighting both the potential benefits and the problematic challenges that AI poses for Indigenous Peoples worldwide.
Among the possible positive uses of AI for Indigenous Peoples are the preservation and revitalization of Indigenous languages, the transmission of traditional knowledge, the management of territorial, historical, and statistical data, the protection and stewardship of territories and natural resources, among other possibilities.
Regarding the obstacles or difficulties that Indigenous Peoples face not only with AI but with digital technologies more generally, the document cites infrastructural limitations, socioeconomic deficits, linguistic and/or cultural barriers, as well as discrimination and the exclusion of Indigenous perspectives from the design and priority-setting processes related to digital technologies.
Additionally, several specific challenges that AI presents to Indigenous peoples and communities are listed, including:
- Fear that the use of digital technologies may alter or cause the disappearance of their cultural practices;
- The protection of knowledge and data sovereignty in the face of the risk of cultural appropriation;
- The lack of control affecting Indigenous Peoples’ sovereignty over their data;
- The absence of adequate legal frameworks to protect intangible cultural heritage.
Digital Divide
The document addresses the issue of structural causes affecting internet access for Indigenous communities—an essential prerequisite for understanding and using AI.
It states that, in various regions of the world, coverage rates vary between 10% and 50%, with difficulties related to electricity connectivity. Consequently, some communities have developed their own community networks, albeit generally with limited reach.
These inequalities, when pertaining to Indigenous Peoples, result from multiple factors: technical, economic, educational, linguistic, political, cultural, among others.
It is noted that even in countries with high levels of development, such as Canada or the Scandinavian nations, these disparities are also evident.
In the vast country of northern America, the Internet Society asserts that while 97% of Canadian urban households have broadband access, this percentage drops to merely 24% in rural Indigenous communities.
Data Sovereignty
Indigenous Peoples must possess the means to create, manage, and control their own databases, their information, and to use it in any manner they deem appropriate.
In this regard, the CARE Principles for Indigenous Data Governance (Collective Benefit, Authority to Control, Responsibility, Ethics), formulated by the Global Indigenous Data Alliance, are particularly noteworthy.
The scope of each principle is as follows:
- Collective Benefit: Indigenous data should benefit the communities, not only researchers, corporations, or States;
- Authority to Control: Indigenous Peoples must retain control over decisions concerning data that pertains to them;
- Responsibility: Data users must acknowledge the cultural, spiritual, and political implications of their use of data;
- Ethics: Indigenous norms must guide data management, including within technologies based on artificial intelligence.
According to the Permanent Forum’s document, these principles constitute an alternative to the Western open-data paradigm, which is often incompatible with Indigenous conceptions of confidentiality, collective responsibility, or knowledge protection.
The document also highlights that, in some countries, criteria for the ownership, control, access, and possession of data have been proposed.
Effective application of such principles or protocols is necessary to prevent digital extractivism, which has been prevalent historically and continues to be so today.
In particular, it is necessary to protect the knowledge of Indigenous Peoples in fields such as ecology, medicine, food production, among others. Digital technology, including AI applied within an appropriate ethical and organizational framework, can be highly useful for preserving and transmitting that knowledge; however, improper use may produce the opposite effect.
The non-consensual extraction of cultural content—narratives, songs, artistic motifs, medicinal knowledge, or sacred toponymy—represents a genuine threat. While this has always existed, the use of AI can amplify it.
The document provides examples from the Amazon region and the northern European boreal region, where the application of AI together with drones and sensors enables rapid alerts to communities in the event of threats to forests, rivers, or hunting grounds, while also improving mapping useful for verifying territorial delimitations.
Internationally, the document includes several examples of protocols developed by Indigenous organizations aimed at regulating research and, more generally, data management and technology use.
Among them:
- Māori protocols (Aotearoa) concerning cultural data (tikanga, whakapapa, mana);
- First Nations protocols (Canada) on research and data ownership, as well as the principles of ownership, control, access, and possession;
- Sámi guidelines regarding their knowledge, archives, and digital representation;
- The CARE principles of the Global Indigenous Data Alliance (GIDA), already cited.
Examples of national principles or guidelines are also provided, such as:
- The principles of ownership, control, access, and possession of the First Nations of Canada;
- The Māori Data Sovereignty Network (Te Mana Raraunga), New Zealand, which imposes obligations of consultation, control, and collective benefit;
- The Draft Guidelines on Sámi Data Governance, aimed at regulating the collection and use of data concerning Sámi lands;
- The principles of ownership, control, access, and possession developed in Canada by the Assembly of First Nations, which establish that the community owns the data, must control its use, may restrict access to it, and may also assume its physical custody through servers, infrastructure development, etc.
Algorithmic Biases
A key issue is the very content of the information available through AI—or, in other words, the bias of algorithms, which may lead to discriminatory decisions and uses of this tool.
As UNESCO states in a recent document, the number of companies and individuals developing these technological tools is not only very small but also concentrated in a few locations (ten thousand individuals in seven countries). This represents a genuine oligopoly of mega-corporations that prioritize the interests of their stakeholder groups.
Within this framework, the inclusion of algorithmic biases in products developed from partial, georeferenced, culturally and commercially predetermined perspectives is an inevitable correlation.
The Permanent Forum’s document asserts that this dynamic contributes to cultural homogenization, a process already reinforced by social media, search engines, and digital industries.
It states that, for Indigenous Peoples, the risk is twofold:
- For their languages and forms of expression, which are often rendered invisible;
- Due to the inclusion of content that reflects hegemonic cultural models.
Data Governance
The document emphasizes that, in several countries, national legal frameworks still do not recognize the collective rights of Indigenous Peoples over their data, whether cultural, linguistic, or environmental.
The problem lies in the fact that, without adequate legal protection, data may be redirected to private actors, sold to third parties, utilized in commercial artificial intelligence models, or incorporated into public policies without any consultation.
Precisely, one objective of the UN Global Digital Compact is to promote an inclusive global framework, essential for multi-stakeholder action aimed at overcoming differences in data and innovation towards an open, free, secure, human-centered digital future grounded in universal human rights.
To this end, it proposes closing digital divides; expanding inclusion in the digital economy; fostering an inclusive, open, and secure digital space; promoting responsible and equitable data management; and improving international governance of artificial intelligence so that it benefits humanity as a whole.
The reference to the respect and promotion of human rights is cross-cutting, including a commitment to uphold international human rights law throughout the lifecycle of digital and emerging technologies, combating all forms of discrimination.
Recommendations
As is customary in such documents, the text concludes with a set of recommendations addressed to various actors (governments, corporations, researchers, non-governmental organizations, and Indigenous Peoples themselves).
Among other points, the following are recommended to States:
- Structural investments aimed at closing the digital divide within Indigenous Peoples’ territories;
- The creation of Indigenous technology hubs, innovation centers, and digital cooperatives, along with the provision of financial support to these entities;
- The establishment of Indigenous innovation funds to enable communities to develop their own technological solutions;
- Recognition of the right of communities to control the use and management of their data and access based on the CARE principles;
- The creation of public programs for technological co‑development with Indigenous Peoples;
- Mandatory integration by governmental bodies of Indigenous representatives in technological projects that concern them.
Technology companies are recommended to facilitate the accessibility of services for Indigenous communities.
This implies:
- Designs compatible with low-connectivity environments;
- Developing such designs jointly with the communities;
- Adequate linguistic interfaces; and
- The provision of culturally appropriate devices adapted to local realities.
Companies are also called upon to respect the CARE principles concerning ownership, control, access, and possession of data.
Two additional lines of recommendations are highly relevant in this document.
Technology companies are urged to integrate Indigenous persons into the design of models, participation in advisory committees, and collaboration with Indigenous educational institutions to reinforce the capacities of communities and their technical personnel.
Furthermore, companies in the sector are requested to conduct periodic audits in coordination with Indigenous communities to detect and, where appropriate, correct algorithmic biases, particularly linguistic, territorial, and, more generally, cultural biases.