3rd Meeting, 19th Session Expert Mechanism on the Rights of Indigenous Peoples (EMRIP)

Good morning, everyone. On behalf of the EMRIP, I want to warmly welcome all our esteemed panelists and participants, including Indigenous Peoples' representatives, member states, national human rights institutions, United Nations entities, academic institutions, Non-governmental organizations, OHCHR Indigenous Fellows, grantees, and all others joining our discussion today. It's my pleasure to chair this timely and very important panel on the rights of Indigenous peoples and artificial intelligence. As artificial intelligence rapidly transforms societies across the world, it's essential that discussions on these technologies fully incorporate the rights, perspectives, and priorities of Indigenous peoples. AI presents significant opportunities, including for the revitalization of Indigenous languages, the preservation of oral histories and knowledge systems, and the transmission of cultural heritage to future generations. At the same time, it raises profound concerns including algorithmic bias, misrepresentation of indigenous knowledge systems and perspectives, exclusion from decision-making processes, the extraction and use of indigenous peoples' data without their free prior informed consent. And the use of AI-enabled surveillance and profiling practices that may disproportionately affect Indigenous peoples. Moreover, these developments risk creating new forms of digital inequality, colonialism, technological exploitation, and dispossession that can undermine Indigenous peoples' self-determination and data sovereignty. These challenges underscore the need to ensure that AI governance frameworks are firmly grounded in international human rights standards, particularly the UN Declaration on the Rights of Indigenous Peoples. The central message of today's panel is clear: the future of AI must be shaped in a way that respects, protects, and fulfills the rights of Indigenous Peoples. This includes recognition of Indigenous data governance principles and Indigenous Peoples' authority over data concerning their communities, cultures, territories, and knowledge systems. It requires respect for the rights to self-determination, meaningful participation in decision-making, free, prior, and informed consent, as well as control over cultural heritage, traditional knowledge, and Indigenous data sovereignty. These principles must guide the development, deployment, and governance of AI technologies, and inform efforts to advance inclusive, ethical, and rights-based approaches to AI. At the same time, we are witnessing the growing leadership of Indigenous peoples in the area of data governance and digital innovation. Across regions, Indigenous-led initiatives are demonstrating how technology, including AI, can be developed and used in ways that strengthen languages, protect knowledge systems, support responsible stewardship of lands and resources, and advance collective well-being. These experiences remind us that Indigenous Peoples are not only rights holders whose interests must be protected, they are also innovators and partners in shaping ethical, inclusive, and sustainable technological futures.

This—

today's panel provides an opportunity to examine the implications of AI for Indigenous Peoples. Highlight Indigenous-led good practices, and identify pathways to ensure that emerging technologies are governed in a manner consistent with the UN Declaration on the Rights of Indigenous Peoples. The outcomes of today's discussion, including its key recommendations, will inform the Expert Mechanism's annual report to the Human Rights Council. As we begin our discussion, I encourage our panelists to propose actionable recommendations for states, Indigenous peoples, United Nations entities, and other stakeholders to support rights-based and inclusive AI governance. Today's panel brings together Indigenous experts, rights holders, policymakers, and technical practitioners. We have 6 distinguished speakers with us, each bringing a wealth of experience, knowledge, and insight into this important discussion. We're honored that you've accepted our invitation and taken the time to join us today. And I want to thank you and give you a warm welcome once again. And I will start now by giving the floor to Maren Nystad Storslett representing the Sami Parliament in Norway. Madam Storslett is a member of the Sami Parliament's Executive Council and a Sami politician representing the Norwegian Sami National Association. She has professional experience in indigenous governance and policy development within the Sápmi region. Her areas of expertise include protecting, developing Sami languages, as well as digitalization and artificial intelligence. Madam Storslett, you have 7 minutes. Thank you.

Kiitos. Moonen oskallama kreddaari Stina Maria Maren ja Otsun Samedikkiin arkas. Chair, distinguished members of the Expert Mechanism, colleagues, and fellow indigenous peoples. I would like to thank the Expert Mechanism for the invitation to contribute to this important and highly relevant panel. The rapid development of artificial intelligence is transforming our society beyond what we can imagine. For Indigenous peoples, things— this brings both significant opportunities and serious risks. For instance, in areas where human resources with language expertise are limited, AI can serve as a vital tool to ensure our languages remain accessible. AI can also strengthen public services, healthcare, education, and access to justice. It can also offer vital opportunities for development within traditional livelihoods, cultural heritage, and traditions. Today, however, Indigenous peoples remain largely underrepresented in national, regional, and international processes shaping AI regulations, standards, and ethical frameworks. Indigenous peoples must take a leading role in discussions on AI at all levels, including within the UN, with a human rights-based approach grounded firmly in the UN Declaration on the Rights of Indigenous Peoples, particularly the right to self-determination, participation, and the free, prior, and informed consent. And I'm warmly welcome the expert mechanism's timely focus on this issue. Chair, in Norway, the government has launched a national AI initiative. For the Sami Parliament, it is crucial to ensure that we are included in these national efforts so that new language models reflect our realities and benefits our communities. We also see an important role of strengthening collaboration with regional processes, for instance within the European Union and its institutions, to ensure that emerging frameworks do not become restrictive barriers, but instead incorporate safeguards for indigenous rights and self-determination. In Sápmi, we have the Sámi AI Lab at the Sámi University of Applied Sciences. They have already developed vital tools for the Sámi community, despite limited funding. Capacity building and support of Indigenous-led centers of expertise and institutions is highly important to ensure our meaningful participation and capacity on these matters. We also need to strengthen cooperation among Indigenous peoples on AI. We must build alliances and networks between Indigenous experts and institutions to exchange best practice, address shared challenges, and co-develop AI systems. With adequate resources and clear mandates, these institutions can drive technological development on our own terms and in accordance with our priorities. They serve to steward and quality assure AI systems being deployed, preventing misrepresentation, algorithm bias, and further marginalization. As AI systems increasingly depend on large-scale data collection, and many AI models are trained on data collected from open sources, Indigenous peoples must have real influence over how data relating to our languages, culture, people, society, and knowledge systems are collected, governed, and used. Data sovereignty and Indigenous data governance is a neutral extension of the right to self-determination. Also, algorithm bias within AI systems may produce stereotypical and harmful representation in AI-generated outputs. If AI models are not trained on accurate or representative data, they fail to understand our realities. In critical sectors like healthcare, this can result in unreliable tools that provide wrong health assessments. The consequent— the concentration of technological power within a few global corporate actors also raise critical questions whose values and priorities are embedded in these AI systems. Just as we have ethical standards and corporate responsibility frameworks in other major industries, we strongly suggest that something similar should be expected from the tech industry. Furthermore, we must not only ask what AI can do but what it should do. AI is resource-intensive intensive and requires vast amounts of energy. In Sápmi, we already see how large data centers put immersive pressure on our territories, and mining activities related to this development of these data centers threatens indigenous people's lands. This comes on top of other extractive activities, and this forces a conversation about priorities and limits, and we need to be at the table on these discussions. Chair, respect for the rights of Indigenous peoples must apply across the entire life of AI systems. This includes the fundamental principles of free, prior, and informed consent. We must ensure Indigenous peoples' participation in decision-making, the governance of Indigenous data and knowledge, and promote forms for technological development that strengthen our communities. And I would encourage the MREP to consider studies and recommendations on the implementation of UNDRIP in the age of AI, ensuring that AI becomes a tool for empowerment, not further marginalization, so that we are on the forefront of this development. Chair, I look forward to hear other inputs on this agenda today.

Thank you very much, Madam Storslett. Now I'd like to give the floor to Emily Pugin, High Commissioner to Cyprus, former director in Australia's Office for First Nations International Engagement. She's a proud Kombumeri First Nations woman And has served overseas at the Australian Embassy and the Permanent Mission to the United Nations in Vienna and Austria. Madam Pugin, you have 7 minutes.

Thank you, Chair. Kayala, colleagues, which means good morning in the Kombumerri dialect. I would like to begin by thanking the Expert Mechanism for its leadership in advancing the rights of Indigenous Peoples. As digital technologies, including artificial intelligence, rapidly reshape societies, economies, and public decision-making. This is a timely and important discussion. For Australia, the central question is not whether AI will affect Indigenous peoples, but whether it is developed and governed with Indigenous peoples in accordance with their rights, priorities, and knowledge systems. AI can be a powerful tool for opportunity and inclusion when it is shaped by the people and communities that it affects. However, without effective governance, it can also deepen exclusion, reproduce bias, misuse data, and undermine cultural identity. Australia strongly supports a human rights-based approach to AI governance, and in the context of Indigenous peoples, this is guided by 3 priorities: rights-based governance, Indigenous data sovereignty, and meaningful participation. These priorities are underpinned by safeguards, Indigenous-led opportunity, partnership, and international cooperation. AI governance should be grounded in human rights and the United Nations Declaration on the Rights of Indigenous Peoples, guided by self-determination, participation in decision-making, free, prior, and informed consent, and protection of cultural heritage. Indigenous peoples must be active partners and decision makers in how systems are designed, used, assessed, and governed through the life cycle of AI. Indigenous data sovereignty must also be central to human rights respecting AI. Indigenous peoples have the right to govern the collection, ownership, access, interpretation, and use of data relating to their communities, lands, cultures, languages, and knowledge systems. In AI systems, data shapes design, training, outputs, and the assumptions embedded in automated tools. Respecting Indigenous Data Sovereignty is essential to ensuring that AI systems are accurate, culturally appropriate, accountable, and trusted. Australia recognises the importance of strengthening data sovereignty so that First Nations communities retain appropriate control over how their data is accessed and used. Meaningful participation must be enabled through targeted safeguards, digital inclusion, and Indigenous-led opportunity. AI can create real opportunities for Indigenous communities, when it is designed, deployed, and governed responsibly. When Indigenous peoples lead innovation, AI and digital tools can support language revitalization, cultural preservation, land and sea management, environmental monitoring, service access, and economic participation. However, there are also risks to Indigenous peoples. These include algorithmic bias, harmful stereotyping, misappropriation of cultural and intellectual property, use of data without proper consent, and outputs that fail to reflect cultural context, language diversity, or community authority. Addressing these risks is critical to building trust and ensuring that innovation does not come at the expense of rights. As such, Australia supports safeguards that are transparent, accountable, culturally informed, with stronger protections for cultural knowledge and clearer redress where harm occurs. Technology should strengthen, not replace, Indigenous knowledge systems, with communities defining the problem, shaping the solution, and governing the outcomes. Participation also requires practical action to address digital exclusion, which risks compounding existing inequalities and as AI becomes more embedded across economies and societies. Australia is working to address this through digital inclusion, connectivity, skills development, AI literacy, and community-led initiatives, so that Indigenous peoples are protected from harm and empowered to participate in and benefit from the digital economy. Australia's Department of Industry, Science and Resources is leading national work on responsible AI. This includes whole-of-government policy engagement focused on safe, inclusive, and human rights-respecting approaches. Australia's National Action Plan continues to build on our robust legal and regulatory framework, and AI ethics principles are designed to capture opportunity, spread benefits, and ensure the responsible design, development, and use of AI systems. These principles support Australia's efforts to harness the potential of AI safely and responsibly. Australia will also establish the Australian AI Safety Institute to monitor and test AI systems, provide timely insights on emerging risks, and support fit-for-purpose protections that keep pace with technological change. The Australian Government is also committed to legislating a digital duty of care to the Online Safety Act, which will require online service providers to undertake thorough and regular risk assessments and implement effective mitigation strategies to address new and emerging AI-related risks and harms. Across these priorities, partnership is fundamental. Australia emphasises sustained engagement so that Indigenous peoples help shape AI governance frameworks, safeguards, policies, and programs, including through co-design and support for Indigenous-led solutions that reflect local priorities and cultural authority. Effective AI governance at all levels requires that First Nations communities have a meaningful voice in shaping their digital future. Not only through access to technology, but also through sharing in the benefits that it creates. Effective governance cannot be developed for Indigenous peoples without their perspectives, knowledge, and leadership. In closing, Chair, international cooperation is critical. AI is a global technology, and its challenges cannot be addressed by any one country alone. Australia is committed to working with partners to promote inclusive AI governance, share approaches to Indigenous data governance, and ensure global standards reflect diversity, cultural context, and human rights. Forums such as EMRIP have an important role in ensuring Indigenous perspectives are reflected in global discussions on AI. Australia reaffirms its commitment to digital transformation that is safe, inclusive, human rights respecting, and empowering for Indigenous peoples. The measure of responsible AI will be whether it is developed with indigenous peoples, respects their rights and knowledge, and delivers benefits defined by their communities. Thank you.

Спасибо, Пьюжин. Спасибо большое.

Thank you very much, Madam Pugin, for that presentation. Next speaker is Vasily Nemychkin, Director of International Research Center of He is a professor of Finno-Ugric Studies at the Mordovia State University. He is an academic and legal scholar specializing in Finno-Ugric Studies, indigenous peoples' rights, and civic diplomacy, with experience contributing to research on international cooperation and the human rights frameworks. You have the floor, Mr. Niemiec, in 7 minutes.

Thank you.

Thank you very much, Madam Chair. Good morning, colleagues, experts. Thank you to the Emirate for the invitation, allowing me to speak on such a crucial theme. AI is an instrument which can provide benefits, but can also— it can do good, but it can also pose threats to Indigenous peoples. AI opens for Indigenous peoples broad opportunities for revitalizing languages, economic development, preserving traditional knowledge, environmental monitoring. However, at the same time, AI systems raise new kinds of ethical problems— illegitimate use of traditional knowledge, blurring the identities of indigenous peoples. AI systems can also heighten the negative prejudices and worsen existing forms of stereotypes and prejudices towards Indigenous peoples. The goal today is to find a model where AI technologies can serve the interests of Indigenous peoples and not the other way around. Currently, in many countries, there's active development of legislative regulation of AI, but approaches vary significantly. In Russia, AI legislation is still developing in Russia. There is a national AI development strategy running up to 2030 that was approved by presidential decree. Last week in the Russian Parliament, there was adoption of a bill on AI legal regulation. It's about developing framework approaches to AI development. That document regulates issues of lawful use of the results of intellectual activities. One of the norms contained is to mark all audiovisual materials created with the help of neural networks. This helps protect from hidden manipulation and discriminatory algorithms. Colleagues, as part of today's discussion, I want to draw your attention to collective study that was developed by organizations of indigenous peoples of Russia in support of an EMRIP survey on the right to data. It's called Indigenous Peoples' Disaggregated Data and Access to Information. That study highlights practices and priorities for indigenous peoples in Russia when it comes to digitalization, including in the sphere of AI. Languages are a key area where AI opens significant opportunities for indigenous peoples in Russia. I'll give you a few examples from our sociocultural region. In Russia, you have a project called Languages of the Peoples of Russia. It's implemented by the IT company Yandex together with the organization, the Home of Russian Peoples, with the support of the Federal Agency on Nationalities. 21 indigenous languages are already accessible on Yandex. 6 languages are available for word recognition and synthesis. In other words, it helps neuronets analyze sound patterns and transform them into words, and also analyze text and then place a human voice with the language of the indigenous language. The plan is that very soon 25 indigenous languages will be available on those services. Now, of course, Similar projects and initiatives on digitalization of languages have already been implemented several years ago in a number of countries, but what's interesting is that in Russia, such initiatives have become a priority for state linguistic policy and have been implemented systemically given the large number of languages in Russia. According to experts, there's probably around With the support of regional associations, we're developing a project with a system of AI called SmartCharm. That's the name, SmartCharm. And it is to help herders using internet and with a satellite connection, and then the person can monitor herding and using a drone quadcopter. Quadcopter, he can monitor direction of deer, use services of telemedicine, et cetera. Another example I'd like to draw your attention to was recently presented by linguistic activists during the 3rd International Summit, World Treasures of Native Languages recently. It's an expand— it's an extension for the Ethnotek browser which automatically translates voice and video tracks into indigenous languages using AI. It allows immediately, immediately without dubbing to maintain the voice intonation and watch videos in the native language of indigenous languages. Currently, this can be done in 15 languages, including in Mardo, Mokshansk, Erzyansk, in Vodsk language. That's one of the small-numbered indigenous peoples. Now, these examples are to be welcomed and supported. We should point out that for practical embodiment of such ideas, significant investment is required and a relevant level of expertise. Colleagues, I'd like to draw your attention also to the fact that AI is a sphere of international competition. AI development has heightened competition between states, between companies, At the same time, in many developing countries, there's an absence of relevant infrastructure. There's not so much in the way of technical opportunities and/or investment that's needed to properly benefit from and implement AI. Such countries therefore often depend on technologies from outside. Separately, there's the need to remove artificial barriers that hinder access for indigenous peoples to new technologies, to protecting their sovereignty and their intellectual property rights. Unfortunately, what we see is that unilateral coercive measures can be used as levers of economic and political pressure. Indigenous organizations in Russia have repeatedly stated at UN fora and called on mandate holders to develop guiding principles of due due diligence to prevent or minimize harm to Indigenous Peoples. In conclusion, I'd like to call on UN mechanisms, specialized agencies in the UN system, to develop an international document. This could be a strategy or a global plan of action that would take into account the rights of Indigenous Peoples when using AI technologies. And also to develop a mechanism for— a complaints mechanism for violations of rights using AI. Undoubtedly, strict ethical regulation is required to avoid illegitimate use of the sensitive knowledge of Indigenous Peoples. And a starting point here could be the ethical principles of AI use that were developed by UNESCO. Back in 2021. Colleagues, AI is not just a technology, it's a reflection of our own values. If we want to build a fair, just future, we need to guarantee that AI can develop inclusively, ethically, respecting the rights and interests of Indigenous peoples. I thank you.

Thank you very much for that statement. Next, I would like to invite Noelani Arista, Director of Institute for Indigenous Research and Knowledge, Faculty of Arts at McGill University. Ms. Arista is an indigenous historian of Hawaii, an expert on the use of artificial intelligence and machine learning to support indigenous language sources, and to increase community access to ʻAike Hawaiʻi through indigenous-centered approaches to knowledge organization. Madam Arista, you have the floor, 7 minutes.

Thank you. Aloha mai kākou. Aloha to the indigenous people of this land and from around the world. Mahalo to the members of the expert mechanism for the invitation to contribute to this important discussion on artificial intelligence and its impacts on indigenous communities. I have 25 years of training and experience in Hawaiian language historical research, as well as textual and archival organization of primary sources from Hawaiʻi, arguably the largest indigenous language textual archive in Native North America, and the Moananuiākea. My training with native speakers was kinesthetic, aural, and oral, and our elders taught us that the greatest threat to our knowledges was forgetting. But they could not foresee the digital present, inundated as we are by a world of words, sound bites, images, and video snippets rendered through social media and generated by AI. At present, The greatest threat to Indigenous knowledges is distortion and the cession of community control to digital mediums. The term data sovereignty is often used. I have heard it in multinational Indigenous projects I work with, but there is little agreement on what the term means, much less the mechanisms or processes that should be codified to ensure that Indigenous knowledges are protected and persist for our futurity. In my work, I have been asked to organize archival collections for greater access, build standalone digital repositories that are open-sourced, and have consulted on Indigenous community projects regarding best practices for AI integration into the management of sensitive community datasets. A singular challenge is that there are no local, regional, or governmental organizations tasked with reaching out and providing education for Indigenous communities on the promise and possible dangers of AI and data organization. Here I would highlight the first of several challenges to traditional knowledge systems: the transformation of TKS into data. How do we best mobilize machine learning and AI applications to support the social relations governed by behavioral norms and community-based ethical standards that transcend a single generation. For most of us, knowledge is not data. Its foremost use is not extraction for wealth generation. It is neither alienable nor property. Put simply, how do we preserve correct knowledge and create supports through AI that do not sacrifice lived relations or violate or devalue longstanding ethical norms and protocols? How do we do so while protecting our lands and waters? Do we or anyone else on the planet have a right to refusal? At present, there are many initiatives to use AI for language reclamation. In my experience, however, many of these teams are led by computer scientists and linguists with few fluent language speakers or traditional knowledge experts shaping the trajectory of work. While governments promote optimistic plans promising AI for all, there are few Indigenous people at the table who are trained in both computer science and Indigenous knowledges and languages, necessitating conscientious and respectful collaboration. Our current AI boom has the energy of a nascent gold rush, and the people who supply funds or technological savvy are the ones who drive decision-making content creation, and delivery. Far too often, project leaders are unfamiliar with our languages, the cultural protocols governing the care and transmission of knowledge, and have not considered the socially corrosive effects of new technologies. Due diligence must be pursued to understand people's credentials and authority so that identity doesn't supplant ike, or traditional knowledge, and community protocols and expertise are honoured. 2 years ago, my colleagues and I started an Indigenous language revitalization program at McGill University, granting both master's and doctoral degrees, as well as an iterative lab where students from history, law, linguistics, and library and computer science may collaborate on various research projects that are dedicated to bringing language and knowledge back to communities. Here, language is not treated as an isolate, but rather as an integral part of putting people back into good relations with one another and giving life to healthy communities. So far, student-led collaborative projects include a digital dictionary co-created with elders in the Swampy Cree dialect, producing machine-read scans of Māori language newspapers, and new standards for language vitality in Ganyengehaga communities. Upcoming projects will map traditional place names, gather historical and contemporary land stewardship practices, and monitor land and environmental transformation over time. We are working collaboratively with students and professors to monitor the legal landscape with respect to our rights and protections. While there are cases wherein a government chooses adoption of the stricter UNDRIP standards into their domestic legislation, Canada, nor the United States has a centralized Indigenous data sovereignty regulatory framework. In the absence of formal legislation, nonprofit organizations and other interest groups have filled the gap, creating several Indigenous data privacy standards that establish more Indigenous-specific norms to govern research conducted on Indigenous communities. While some— sorry— of these standards are recognized and supported by the respective federal governments for internal purposes, they are not formal legislation and therefore not strictly binding. Rather, the intent is setting norms that support Indigenous data sovereignty within government agencies, research spaces, and other third-party organizations. As these standards are guidelines rather than protections with the backing of law, Indigenous stewardship, control, access, and possession of traditional knowledge systems remains at risk from the highly competitive, fast-moving, and unfettered development and implementation of AI systems globally. Therefore, in accord with Article 13 of the UN Declaration of Rights on Indigenous Peoples on the right to revitalize, use, develop, and transmit to future generations our languages, oral traditions, writing systems, and literatures, and Article 14 and 16 on our right to establish educational systems and media in our languages. We call on international bodies and EMRIP to recognize Indigenous expertise and ability to contribute to the development of ethical AI systems and to apply pressure to individual nations to more strongly regulate the entities within and between their jurisdictions in the interest of meaningful protection of Indigenous rights in AI development and the protection of Indigenous peoples' lands, seas and waters. E ola.

Thank you very much for that statement. I'd like to invite Olivier Elias, Program Coordinator at the United Nations International Telecommunication Union. Mr. Elias has worked on integrating human rights considerations into digital governance and technical standards with a focus on artificial intelligence and inclusive digital transformation. Mr. Alias, you have the floor.

Madam Chairperson, distinguished experts, Excellencies, dear colleagues, thank you very much for having invited me to speak on behalf of the International Telecommunications Union. This is a United Nations specialized agency for digital technologies. Today, I'd like to develop 3 points. Connectivity is not yet a right that has fully been acquired by Indigenous Peoples. Artificial intelligence may either exacerbate that exclusion or contribute to remedying it, depending on how it's conceived. And technical standards— these rules which are largely invisible but which determine the behavior of technologies— are a legitimate and thus far underexploited space for Indigenous participation. Let me turn first of all to connectivity then. Today, 2.2 billion people remain unconnected. Indigenous peoples, over 476 million individuals over some 90 countries with around 5,000 different cultures and languages, are structurally overrepresented in this non-connectivity. Thank you. This isn't mere geographical chance. This is a reflection of ancient, old, long-standing patterns of dispossession, removal, and investment in infrastructure which have constantly bypassed Indigenous territories. Where connectivity is lacking, we also see a lack of more recent digital economy tools such as cloud services, AI, and harmonization alignment bodies which discreetly shape these technologies. That is why the ITU is supporting connectivity headed by Indigenous Peoples themselves, and not just simply connectivity that is delivered to Indigenous communities. This is a commitment our member states have formally mandated for over 2 decades now, and this has been recently reaffirmed in Resolution 46 that was adopted last year at the World Telecommunications Development Conference in Baku. This was also based on Resolution 184 adopted by our Plenipot Conference in 2022. A training program that we've been carrying out with regional partners in 19 countries in Latin America have enabled the NASA, uh, Garifuna, Quechua, Huichol, Digá, Kuti, Qalchaki, and Guarani communities to build, exploit, and maintain their own networks. Around 120 individuals, including— and 4 out of 10 are women— have then trained thousands of other people. The lesson that we are learning, and this is the first question, really, is not just to know whether a community is connected, but under what conditions it is connected. Yes. Let me turn more now to AI. This is increasingly relevant here. AI systems are largely trained on a handful of dominant languages, and the assessments presented this year at the WISIS Forum suggest that up to 4 billion people whose mother tongues, cultures, and local knowledge remain underrepresented in the training data they would get very unreliable results or distorted results from AI systems that are now incorporated into public services, into education, and to daily search engines. For Indigenous languages, many of them already threatened after decades of active suppression, and this is not just a simple technical precision problem, it's additional pressure and it's more silent pressure towards deletion. However, this same technology under Indigenous control can produce the opposite effect. in New Zealand. Te Hiku Media have developed an AI tool aiming to collect and preserve oral recordings in te reo Māori. And they've placed that recently collected data under a license, which is— which is held— the governance of which is held within the community and expressly prohibits its use for the purposes of surveillance or discrimination. As one of the designers has said, data is the new earth, and people who have seen territorial dispossession understand exactly why data sovereignty, uh, could— does not— is not something that needs to be thought about on the second— as a secondary issue. This model needs to be further widespread, not just AI applied to Indigenous peoples, but AI governed by them on infrastructure and according to licensing conditions that they control. The ITU Resolution on Multilingualism commits our members to supporting linguistic diversity in ICT and the task now is to expand this commitment in the era of AI. Let me turn on to my third point, and that's technical standards. Member states of the ITU adopted at our last Global Assembly on Telecommunications in 2024 a resolution on our standards work in AI regarding reliability, and our members are currently looking at the simple guidelines which are human-focused, aiming at Identifying, from the beginning of the process of drafting standards, risks linked to human rights before the systems are built, rather than after they've caused harm. And it is precisely at that point that the sovereignty of all Indigenous data, linguistic rights and language rights, and the principle of free, prior, informed consent must be incorporated into the choices of technical concepts, and not just added in years later following a dispute or as part of a policy. Indigenous Peoples have their place in this room and in that room, within the study commissions at the ITU, and in the broader ensemble of the AI governance process, which is building and being shaped. And for example, in the most— in the first UN Global Dialogue on AI Governance that was held in Geneva last week. Madam Chair, the ITU isn't claiming to speak on behalf of Indigenous Peoples, but we are committed to working on 3 specific elements: continuing to work with community networks conducted by Indigenous Peoples, supporting AI in the service of a common good, which includes Indigenous communities, and encouraging Indigenous participation in our digital technical standards processes in line with applicable provisions governing the members of the ITU. Technology is not at the heart of injustices experienced by Indigenous peoples, but if it is devised without caution, it can exacerbate them. It's, um, build If built within a genuine spirit of partnership, it can help Indigenous Peoples to do what they've always done, to preserve their languages, their knowledge, and their self-determination in their own conditions. Thank you

very much for that statement. And lastly, I would like to invite our last speaker, Fayda Dulmarpa, representative of the Indigenous Peoples Caucus of the Global Alliance of National Human Rights Institutions, GANHRI. Ms. Dulmarpa is a Commissioner at the Philippines Human Rights Commission. As a focal point for gender, she leads the Commission's efforts to advance women's human rights. A strong advocate for Indigenous peoples and other marginalized groups, she champions inclusive rights-based approaches that promote dignity, equality, and social justice for all. Ms. Dumarpa, you have 7 minutes, madam. Assalamualaikum,

distinguished members of EMRIP, colleagues, and participants. It is an honor to join you today on behalf of the Global Alliance of National Human Rights Institutions, or GANRI. We thank the Expert Mechanism for the invitation to contribute to this important discussion on Indigenous Peoples and artificial intelligence. Artificial intelligence is already shaping decisions that affect Indigenous peoples' lives in profound ways, from access to public services, land and environmental management, health diagnostics, education tools, to policing and border governance. Yet these systems are often developed without the meaningful participation of Indigenous peoples, and in many cases without adequate safeguards to protect their rights, data, knowledge systems. From a human rights perspective, this raises both urgent concerns and important possibilities. On the one hand, we are witnessing well-documented risks. First, algorithmic bias and discrimination. AI systems are frequently trained on datasets and underrepresent or misrepresent Indigenous peoples, their languages, identities, and realities. This can result in exclusion from services, distorted decision-making, or discriminatory profiling. In some contexts, automated systems reinforce historical inequalities rather than correcting them. This growing use of facial recognition technologies further illustrates these risks. Second, data sovereignty and lack of control over Indigenous knowledge. AI development often involves large-scale data extraction, including cultural, linguistic, environmental, and even sacred knowledge. Without free, prior, and informed consent, this can amount to a form of digital dispossession where Indigenous peoples lose control over how their knowledge is used, stored, and monetized. Third, surveillance and misuse of technologies. AI-powered surveillance tools, including facial recognition and predictive analytics, have in some cases been used in ways that disproportionately affect Indigenous communities, particularly in context of land defense and environmental protection. For example, the Australian Human Rights Commission has highlighted concerns regarding the use of emerging AI-enabled policing and criminal justice technologies, including the deployment of real-time face— facial recognition technologies and proactive policing tools involving drones and smart tracking in areas with significant Aboriginal and Torres Strait Islander populations. The Commission has also raised concerns about AI-driven security risk assessment tools used in places of detention, which may classify individuals according to perceived risks of escape or violence. Fourth, digital exclusion and inequality of participation. Many Indigenous communities continue to face structural barriers to digital access, infrastructure, and capacity. This means they are often absent from the design, governance, and oversight of AI systems that nonetheless affect them directly. In the Philippines, for example, Indigenous peoples have expressed that they find it challenging to access technology not only due to lack of connectivity but also due to the lack of knowledge on technological use. The growing reliance of social and knowledge systems on AI and digital access has also impacted the transfer and exercise of indigenous knowledge and practices, as existing indigenous systems are not digital to begin with. Fourth, the expansion of artificial intelligence carries significant environmental costs. The data centers and infrastructures required to develop and operate AI systems consume substantial amounts of energy, water, and land. In the context— in the context of climate and environmental crisis, the rapid growth of AI may exacerbate existing pressures on ecosystems and natural resources. These impacts can be particularly significant for Indigenous peoples, whose lands and territories are often affected by large-scale infrastructure projects, resource extraction, and environmental degradation. The Australian Human Rights Commission has also drawn attention to proposed AI data centers that may place additional pressure on Indigenous lands and water resources, raising concerns about what has been described as the digital colonization of land and natural resources where Indigenous peoples are not meaningfully consulted or able to exercise free, prior, and informed consent. At the same time, Ganry believes it is important not to view AI only through the lens of risk. We are already seeing examples where AI and digital tools are being used to support Indigenous languages revitalization, including through translation models, voice recognition tools, and digital archives co-developed with Indigenous communities. There is also growing potential for AI to improve access to public services, including health diagnostics in remote areas, early warning systems for climate-related risks, and educational tools adapted to local languages and contexts. The question, therefore, is not whether AI should be used in relation to Indigenous peoples, but rather how it is governed, who controls it, and whose values it reflects. In this regard, national human rights institutions have a critical role to play. Within GANRI, this engagement is reflected institutionally through the GANRI Caucus on Indigenous Peoples established in 2024, which provides a dedicated space for coordination peer exchange, and strengthened advocacy on Indigenous Peoples' rights across regions. NHRI are uniquely positioned between states, Indigenous Peoples' representative institutions, civil society, and international frameworks. They can help ensure that AI governance is aligned with human rights standards and that Indigenous Peoples are not left out of digital transformation processes. Gannery would like to highlight several areas where action is particularly urgent. First, the need to ensure meaningful participation and free, prior, and informed consent in AI-related processes. Indigenous peoples must not only be consulted over after technologies are developed, but be involved from the earliest stages of design, data collection, and governance. Second, the importance of recognizing and operationalizing Indigenous data sovereignty. This includes respecting Indigenous governance over data, ensuring culturally appropriate data stewardship frameworks, and preventive and unauthorized use of Indigenous knowledge in AI training systems. Third, the need for rights-based AI regulation and impact assessments. States and private actors should be encouraged to carry out human rights impact assessment that explicitly consider Indigenous peoples' rights including potential impact on culture, land and territories, language, and collective rights. Fourth, the strengthening of accountability mechanisms and remedy pathways. Where AI systems result in harm, Indigenous Peoples must have accessible mechanisms to challenge decisions, seek redress, and obtain remedies that are effective and culturally appropriate. And fifth, investment in capacity building for the inclusion of Indigenous Peoples in digital systems. Indigenous communities, including Indigenous youth and women, should be supported by actively participating in AI development, including through education, technical training, and support for Indigenous-led innovation. Thank you.

Thank you very much for that statement. Now I'd like to I'd like to address my colleagues, distinguished members of the mechanism. At this stage, they might like to make their comments.

Thank you.

Well, now we'll move to the list of speakers. First speaker, I'd like to give the floor to Guatemala.

Good afternoon. Good morning.

Thank you, Madam Chair. Guatemala recognizes that AI is defining multiple spheres of our societies and creating new opportunities to promote sustainable development, strengthen institutional capacities, and improve provision of public services. Swift evolution of technology makes it necessary to review governance frameworks that enable us to optimize the benefits and properly manage its risks to ensure that development responds to the public interest, persons'— people's well-being, and respect for human rights. As a pluricultural, multilingual, and multiethnic country, Guatemala believes that discussions over AI should incorporate as a priority the perspectives and rights of indigenous peoples. This means recognizing 3 fundamental elements at the very least: cultural diversity, linguistic diversity, and scientific knowledge of indigenous peoples. Those elements should be part of building governance of new technologies and be considered during dialogue processes, policymaking, and decision-making. Indigenous peoples are an essential part of economic, social, and political life. As well as the cultural, linguistic, and knowledge richness that characterizes Guatemala. Our languages, practices, and knowledge systems constitute valuable expressions of human diversity and are part of the cultural intellectual heritage that enriches our societies. Guatemala believes that AI governance should recognize and value Existence of multiple knowledge systems, promoting spaces for dialogue and cooperation that contribute to developing relevant solutions to different contexts and realities. Thank you.

Пасибо вам за вашу

Um, thank you, Madam Chair. And since this is my first time taking the floor, I'd like to congratulate the new chair of EMREP warmly. The rapid development and deployment of artificial intelligence technologies have brought both significant opportunities and serious risks for Indigenous peoples. On one hand, Artificial intelligence presents an opportunity to support Indigenous language revitalization, the maintenance of oral histories and cultural practices, and the transmission of Indigenous knowledge to future generations. It can also improve access to education, health information, and public services when developed and used in ways that respect Indigenous is people's rights. However, the extraction and use of Indigenous knowledge without free, prior, and informed consent, algorithmic bias, the reproduction of harmful stereotypes, the exclusion of Indigenous knowledge systems and perspectives, and the use of artificial intelligence-enabled surveillance, profiling, and law enforcement practices that may disproportionately affect Indigenous peoples present potential risks. In addition to this, these technologies also come with environmental cost. The data centers that power them consume vast amounts of energy, water, and minerals, resources that are often extracted from or developed on Indigenous peoples' lands. Without their meaningful participation or consent, with potential serious consequences for their rights, livelihoods, and environments. Therefore, during the 25th session of the Permanent Forum on Indigenous Issues, the Forum urged the United Nations system, member states, and all stakeholders to include a dedicated section on Indigenous Peoples in all artificial intelligence development and deployment frameworks, ensuring the protection of Indigenous peoples' rights, including Indigenous data sovereignty and the safeguarding of Indigenous knowledge. At its 24th session, the Permanent Forum recommended that all UN entities involved in developing, governing, and applying artificial intelligence and digital technologies ensure the meaningful inclusion of Indigenous peoples in decision-making and related processes concerning these technologies for the benefit of Indigenous peoples. It further called for the development of instruments, safeguards, and grant schemes to promote digital equity and prevent harm from and misuse of digital technologies, particularly in relation to Indigenous women and children. Thank you. Koyennami.

Thank you, Madam Taker, for that statement. Next speaker is Congress Mondial Amazigh. Good morning, ladies and gentlemen.

In the world, MSCA Congress, or CEMA, welcomes the Expert Mechanism's landmark study on the Indigenous Peoples and artificial intelligence and their right to access data, which rightly recognizes that Indigenous data are not merely technical resources but also Indigenous Peoples' inherent rights to self-governance self-determination, and cultural survival. In this context, CEIMA, the World Amazigh Congress, is pleased to announce today that the development and forthcoming launch of the CEIMA Indigenous Smart Governance Platform using its in-house AI expertise. This initiative has been designed as a practical smart governance platform that supports not only the Amazigh nation but also all Indigenous institutions in exercising their rights to self-governance, data governance, and sustainable development. The platform can provide Indigenous organizations with various smart tools for land protection, legal support, environmental monitoring, institutional capacity building, multilingual knowledge management, and AI-assisted decision support. For example, if an Indigenous community have a particular issue, they, they need— they only need to upload their evidence and select options for workflow related to their issue. Then the system will automatically produce content like legal dossier or media content. This AI-based platform was shortlisted as a finalist for the, the best AI research paper at the AI Awards during the last April's event in April. The World Amazigh looks forward to working with the expert mechanism, and if you would like to know more about it, just please get in touch. Thank you very much.

Thank you for that statement. The next speaker is the Russian Federation.

Thank you.

Madam Chairperson, we share the concern on the review of AI looked at through data cybersecurity or benefits for companies to the detriment of the human dimension. Russia has rich experience in this area over the first act that was codified over 200 years ago on indigenous However, the growing emergence of AI is bringing new realities. AI is being used in Russia to develop satellite surveillance and to assess the impact of industrial projects on land and diversity.

AI creates new possibilities to revitalize indigenous languages in times of globalization. In Russia, the number of such projects with state and business is growing fast. However, at the same time, there are problems protecting authenticity and sanctity of relevant data, intellectual property, and the status of their carriers, which doesn't fit into patent law. Similar processes are affecting the spheres of biodiversity, food, medicine, and others. The mass culture of AI, which learns from data, is speeding up the blurring of traditional knowledge, and when— and with using process of developing solutions could marginalize the values and way of life of indigenous peoples. In Russia, that we've adopted a code of ethics in the sphere of AI. It separately touches on issues of protecting ethnicities, culture, and language diversity, traditional values, and lifestyle, lives of peoples of Russia. However, in the context of protection of indigenous rights, it's necessary for international efforts to expand general awareness and to develop universal approaches that are able for— that are open for subsequent adaptation at national levels. Thank you.

Thank you. The next speaker, Kimberley Aboriginal Women's Council.

Chair, I speak on behalf of First Nation women from the Kimberley region of Australia. As the world rapidly embraces artificial intelligence, we remind this forum that Indigenous peoples have always held and transmitted another form of intelligence, our ancestral intelligence. Our ancestral intelligence intelligence is carried through collective wisdom passed between generations. As AI systems expand, we face both opportunity and risk. Emerging technologies may support language revitalization and community well-being. However, without safeguards, AI also risks becoming a form of digital colonization. We are increasingly concerned that AI systems are reproducing our artworks without respect for cultural protocols, Sacred symbols and traditional artistic styles are being copied, modified, and commercialized by AI, eroding cultural integrity and depriving communities of the economic and cultural benefits that rightfully belong to them. The rights affirmed in UNDRIP must guide all AI governance. We must have authority over how our data, knowledge, and cultural expressions are collected and incorporated into AI systems. For us, this is particularly important. Our voices must be present in the design, governance, and regulation of emerging technologies. Chair, as frameworks for AI are developed, they must also commit to protecting and respecting ancestral intelligence. The future of technology must not come at the expense of the world's oldest living knowledge systems.

Thank you.

The challenge before us is not simply how to regulate AI, but how to ensure it strengthens rather than diminishes Indigenous peoples' rights, cultures, knowledge systems, and futures. Thank you, Chair.

Thank you very much.

Next speaker, the EU.

Distinguished members of the expert mechanism, distinguish Indigenous peoples' representatives. The European Union reaffirms its unwavering commitment to the rights of Indigenous peoples, as affirmed in the United Nations Declaration on the Rights of Indigenous Peoples and international human rights law. Artificial intelligence offers important opportunities to support the revitalization of Indigenous languages, preserve cultural heritage, and strengthen the transmission of Indigenous knowledge However, AI must not reinforce existing inequalities or enable the misuse of indigenous knowledge, data, or cultural expressions. The EU believes that AI should be safe, secure, trustworthy, human-centric, and developed in full respect for human rights. This includes ensuring the meaningful participation of indigenous peoples in the design, development, and governance of of AI systems that affect them while respecting the principle of free, prior, and informed consent and Indigenous data governance. We welcome the growing number of Indigenous-led initiatives demonstrating how AI can contribute to language revitalization, education, and environmental stewardship. Such initiatives show that innovation and respect for Indigenous rights can and should go hand in hand. Distinguished members of the expert mechanism, how can states best ensure that indigenous peoples are genuine partners in shaping AI frameworks while safeguarding indigenous data sovereignty and enabling communities to benefit from these technologies? I thank you.

Thank you. Next speaker.

Inuit organization.

Thank you, Chair.

My name is Julia Eggeville, and I am an assistant professor at the Department of Journalism at Elisabethus Haafik, the University of Greenland, and I will speak on behalf of the Inuit Circumpolar Council. ICC welcomes this panel and the attention to the implications of artificial intelligence and the rights of Indigenous peoples. A digital transition and the use of AI must be done with free, prior, and informed consent of Indigenous peoples, respecting Indigenous data sovereignty. Far too often, Indigenous knowledge has been extracted and used outside its cultural context without including Indigenous knowledge holders in the process. It is critical when we discuss movements towards the use of AI that Inuit are fully engaged in addressing the risks and threats identified. A helpful example comes from Kalaallit Nunat, Greenland. A local news outlet has developed, developed its own AI translation tool, which is trained on news articles in our indigenous language, Kalaallitsoot. It has been developed to reflect and strengthen the actual language and context of the community. In our journalism program, the new curriculum introduced this year stipulates that students must be taught how to use AI. It is crucial that this instruction not only focus on using AI tools, but also teach students to approach them with both ethical and practical critical thinking. We thus recommend sharing knowledge on rights-based approaches in teaching indigenous students in AI tools. Thank you.

Thank you. And I give the floor to social organization Saving the Eagles from the anti-Mansisk

Distinguished Madam Chair, I represent indigenous peoples of Hungary, bringing together the Hanti, Mansi, and Nenzi people. AI development requires not just technological solutions but new models for cooperation. Technological companies and representatives of indigenous institutions must together shape digital linguistic resources guaranteeing quality data, respecting cultural specificity, and upholding the rights of indigenous peoples. We already see that AI can be an effective instrument for preserving languages in Yugra with creating digital linguistic repositories. We're developing linguistic services. This experience shows that modern technologies can not only document languages but also expand the space for their everyday utilization. When developing AI systems, it's important to take into account not just language data, but the cultural context, traditional means of transmitting knowledge, and worldview of Indigenous Peoples. Without that, even technologically accurate solutions could distort the sense of cultural heritage. Special attention should be given to training specialists from Indigenous Peoples in digital technologies, linguistics, and AI. This will help the communities themselves to participate in creating and developing modern linguistic technologies. AI shouldn't replace carriers of languages or teachers or the protectors of traditional knowledge. Its goal is to strengthen opportunities for human— for people and not substitute living linguistic cultural environment. We recommend the EMRIP to continue thematic discussions on AI and rights of Indigenous peoples. We call on the UN system institutions to develop practical guidelines on ensuring the rights of Indigenous peoples during development, mainstreaming, or implementation of AI technologies. Thank you.

Thank you. And I give the floor to Mexico.

Thank you very much, Madam Chairperson. We thank the panelists for their interesting presentations this morning. Mexico believes that artificial intelligence can be a tool to promote the rights of Indigenous Peoples and the dissemination of Indigenous languages and cultures, as long as it is done under a human rights-based approach and with the CARE principles for the governance of Indigenous data. And that it recognizes that Indigenous Peoples are rights holders and therefore holders of information and the data provided and used. Essentially, what is required is the free, prior, informed consent of Indigenous Peoples in initiatives to develop and to use AI based on the Declaration on the Rights of Indigenous Peoples. Mexico adopted a constitutional reform in 2024 which recognized that Indigenous Peoples are rights holders with legal personality and their own heritage. We are now drafting a general law proposal that will be submitted to a consultation process with the 16,728 communities of the 69 Indigenous Peoples and the Afro-Mexican people of Mexico. In this draft bill, We recognize that Indigenous peoples have the options of establishing and operating their own media, radio, telecommunications, and ICT technologies that are community and Indigenous, as well as proposing, implementing, and accessing training and advanced capacity building in ICTs and AI, which would enable them to update content and with relevant, culturally relevant content. We believe that the principal element around the use and development of AI is the building of national and global effective regulatory systems which would recognize Indigenous people as rights holders, promote their full and effective participation in the development of this technology, and provide ethical support and culturally relevant support based on the Declaration. Thank you.

Thank you for that statement. statement. I give the floor now to International Indian Treaty Council.

Thank you, Madam Chair, and many thanks to the speakers who addressed key challenges related to Indigenous data sovereignty and AI governance, including digital colonialism. We are in an urgent situation regarding the infrastructure being built on a massive scale to power AI technology. Data centers are being constructed across the globe, impacting land, water, and energy, and exacerbating the climate crisis. Every major tech company, including Amazon, Google, Meta, Apple, and OpenAI, which fuels ChatGPT, are racing to build multi-million dollar data centers. Just last week, it was announced that the government of Alberta will partner with Meta to build a $13 million AI centre near Edmonton, set to use more than 7 billion litres of water each year, and without the free, prior, and informed consent of the treaty nations that will be impacted. Extensive data centres are being proposed and constructed on and near other Indigenous territories and demand access to local water resources as well as tremendous amounts of of electricity. The energy needs of the data centers are often so large they require new power plants to supply their demand. We must also acknowledge the health impacts, especially on our women and children. These centers release dangerous fumes that include fine particulates and penetrate lungs and can enter your bloodstreams. These pollutants can trigger and worsen health issues including asthma, and heart attacks, and in recent reports shared by National Health Institutes, data centers have been linked to miscarriages and rare cancers. We call on the MRIP to include this topic in a future study with a special focus on the use of water needed for these centers and support Indigenous peoples around the world calling for a moratorium on the creation of data centers until the dangers to our health and lands and territories can be addressed. Many thanks.

Thank you very much for that statement. Next speaker, National Indigenous Women's Association, Nepal. Next speaker, New Zealand.

Thank you, Chair, and thanks to the panelists for their interesting presentations. Distinguished delegates, the Māori economy is a significant contributor to New Zealand's current and future economic prosperity across a wide range of economic sectors, including in digital technology and AI. To support the continued growth of the Māori economy, the New Zealand Government released Tō Nui Māori: Going for Growth with Māori in February 2025. This framework sets out a coordinated approach to unlocking the productivity potential and comparative advantages of the Māori economy. While agriculture, forestry, and fishing remain foundational industries that are deeply linked through ancestral ties, Māori economic activity is increasingly expanding in areas such as professional and scientific services, media, renewable energy, and research and development, with many businesses utilising AI to improve efficiency and accessibility of their services and outputs. For example, Māori-led company Carepatron has developed an AI-powered clinical support tool integrated into practice management platforms to deliver improvements and scalability for health providers. The Māori economy is well positioned for continued growth, supported by a growing and increasingly skilled workforce, the potential of Māori-owned land, and the long-term investment approach of Māori businesses and organisations. Through Going for Growth with Māori, the Government is focused on accelerating growth by supporting the productive use of Māori land, increasing Māori participation in infrastructure investment, and helping Māori businesses expand into international markets. Together, these priorities will help unlock economic opportunities, strengthen regional development, and contribute to New Zealand's economic success. Thank you.

Thank you.

Next speaker, Asia Indigenous Peoples Caucus.

Distinguished experts, I'm delivering this statement on behalf of the Asia Indigenous Peoples Caucus. Our languages, oral histories, indigenous knowledge, faces, cultural expressions, and community data are already being scraped, digitized, and compiled without our knowledge, control, nor free, prior, and informed consent. Information already available about Indigenous peoples has often been created by others and can be incomplete, biased, or rooted in colonial perspectives. We are alarmed that governments, academics, and business enterprises are creating datasets from our Indigenous knowledges and languages and making them accessible for artificial intelligence use. FPIC cannot be reduced by simply clicking a box or digitally accepting the terms of agreement many in our communities may not read or even understand. Our knowledge is collective. Decisions about its use must be grounded in our customary laws, our collective decision-making processes, and our Indigenous government systems. We are also deeply concerned by the lack of access to technology that many Indigenous communities face across Asia. Many Indigenous communities do not have reliable internet or basic digital infrastructure. Indigenous women, youth, elders, and persons with disabilities are disproportionately affected. Artificial intelligence cannot be understood as inclusive when Indigenous peoples have been left out of its design, governance, and use. Therefore, we call upon States, the United Nations, and technology companies to respect and protect our collective rights and Indigenous data sovereignty, ensure full and effective participation, and obtain our free, prior, and informed consent in the development and application of artificial intelligence. Further, companies must prevent our knowledge, languages, and cultural expressions from being used and commercialized without our collective consent, and funding must be secured for Indigenous-led digital infrastructure. infrastructure, artificial intelligence literacy, and technologies that our communities have direct control over.

Thank you. Next speaker, Organization Nacional Indígena Colombia.

Not present.

Next speaker is Norway.

Chair, it's a privilege to address EMRIP on artificial intelligence and Indigenous Peoples' rights, an area where technological development and human rights are increasingly interconnected. Artificial intelligence is rapidly transforming how information is produced, shared, and used. It shapes access to knowledge, public services, and decision-making processes. While this creates opportunities, it also raises significant human rights concerns. If AI systems are developed without the inclusion of Indigenous peoples' languages, knowledge, and perspectives, they risk reinforcing exclusion. AI is not neutral if it does not reflect the diversity of the societies it serves. This makes inclusion a matter of fundamental rights. Indigenous peoples must be meaningfully involved in the development, deployment, and governance of AI systems. This is essential to ensure participation, accountability, and respect for self-determination. Indigenous-led initiatives show how AI can strengthen languages and support digital participation on Indigenous peoples' own terms. In Norway, cooperation with the Sami Parliament and Sami institutions such as Divun at the Arctic University of Norway and Sami AI Lab at the Sami University of Applied Sciences highlights that digital participation is a core democratic Rights-based AI governance must respect indigenous data sovereignty, and Emre provides an important forum for addressing this issue collectively. The challenges and opportunities posed by AI are inherently transnational and require strengthened international cooperation to safeguard the rights of indigenous peoples. Thank you. Пасибо вам за вашу вопрос.

Thank you. Next. Thank you.

I am Rajkumar Roth, a Member of Parliament of India. Today I have the privilege of addressing the United Nations on behalf of the indigenous people of India. First of all, I would like to refute the, the claim that has been refuted by, made by the Government of India at the United Nations since the 1880s, that there are no distinct indigenous people in India. I believe the claim is actually incorrect, and I personally reject those assertions. India is home to indigenous community, privately primarily belongs to the Proto-Austro-Indo-Negroid, Dravidian, and Mongolian racial group, which are collectively classified in 705 Scheduled Tribe communities. These communities are recognized as Scheduled Tribes in the Constitution of India. Today, our population is approximately 140 million. Therefore, that it evidences this India's Scheduled Tribe of the country's indigenous peoples. This position was also offered by Justice Markand Karju, former judge of the Supreme Court of Indigenous India, in Kelas versus State of Maharashtra, where he stated the Scheduled Tribe question of about the percent of India's peoples are the descendants of the originally 180 states of India. I am proud to belong to the indigenous community of India, and it is honor to represent my community. Thank you.

Thank you for that statement. Next speaker. Lucy Marcela Suya from the organization Chovenesa. Kiwi. Kiwi.

Muchas gracias por concederme la palabra.

Thank you very much for giving me the floor. I'm Lucy Marcela Clunque. I'm from the Inasa indigenous people from the I'm here thanks to the ancestral lands in Colombia. I'm here thanks to— Yes. AI is transforming the world, but it cannot advance ignoring our rights. For indigenous peoples, land is not a resource. It is our heritage. It is our spirituality. It's our history, our spirituality that we've protected for generations. We're troubled that AI development There's growing exploitation of minerals, consumption of water and energy, and pressure on our territories, often without any ethic. And when a territory is destroyed, not only do you lose biodiversity, but also they break our spiritual relations with Mother Earth. Our ceremonies, sacred places are affected. AI also raises risks to our traditional knowledge, our languages, medicine, cultural practices and ancestral knowledge cannot be misused for training AI systems without our ethic. That consent are just free access data. They're part of our collective identity. Therefore, we urge the United Nations and states to ensure that AI governance should fully incorporate the rights of Indigenous peoples, guarantee our participation and respect our self— free self-determination. Technological innovation should never be built at the cost of our territories, culture, or spirituality. AI should not be converted into a new form of digital colonialism. Thank you.

Thank you. Next speaker is Canada.

Thank you, Chair. Canada welcomes this discussion on the rights of Indigenous peoples and artificial intelligence. This summer, Canada announced the launch of a new artificial intelligence strategy, AI for All, which will guide new legislation, policy, investments, and programs to support the responsible development and use of AI to serve all Canadians. Canada recognizes both the opportunities and risks associated with AI. These technological innovations also raise important human rights considerations. Canada notes and appreciates the 2025 study of the expert mechanism regarding AI and data sovereignty and the application of human rights norms, particularly the considerations about the collection, access, use, and sharing of data by and about Indigenous peoples. With care and intent, artificial intelligence can represent a unique opportunity to support Indigenous language revitalization efforts. However, it is important that we consider protecting Indigenous data sovereignty and advancing Indigenous knowledge and languages with consent. Canada will work with Indigenous peoples to ensure AI tools designed and built domestically will protect Indigenous cultural and linguistic expression and support Indigenous self-determination over how AI is built and used in Indigenous contexts. As AI increasingly influences decision-making, Canada affirms that Indigenous data sovereignty must extend to data, algorithms, and systems. AI can support the preservation of Indigenous languages, knowledge, and cultural practices, but only when developed with free, prior, and informed consent. Without these safeguards, AI risks reproducing historical exclusion, bias, and inequity, reinforcing colonial narratives rather than advancing Indigenous rights and self-determination. Thank you.

Thank you. Next speaker. The organization, web community.

Distinguished Chair, I'm a representative of indigenous peoples of the northwest of Russia, the Veps people, and we're a grantee of the UN Voluntary Fund for Indigenous Peoples. I thank the fund for being able to take part in the work of the expert mechanism. The Veps community welcomes the attention of the experts on issues of Indigenous peoples in the age of development of AI. The headlong development of AI is creating new opportunities to create culture and language of Indigenous peoples, but also raises major concerns, including amongst Indigenous communities. AI can help protect cultural and linguistic diversity only if Indigenous peoples themselves are not just an object of digitalization, but rather full-fledged participants in the creation, management, and implementation of said technologies. We urge member states, academia, and technological companies to expand their support to projects to expand digital dictionary repositories, machine translation, with the leading role played by indigenous peoples themselves. Without systemic forming of language data, many indigenous people will not have access to the AI technologies. Development of digital technologies and linguistic resources is becoming an important part of linguistic policy. Digital technologies must create conditions not just to document languages, but also for their everyday use in education, communication, and digital services. Without the direct participation of language holders, it's not possible to guarantee the quality of language models, their cultural authenticity, and correct interpretation of traditional knowledge. We recommend to the EMRIP to prepare practical recommendations on application of AI to protect languages, traditional knowledge, and cultural heritage of Indigenous peoples, including issues of Indigenous peoples' data management and their participation in development. Thank you.

Next speaker. African Indigenous Foundation for Energy and Sustainable Development.

Thank you, Madam Chair, and congratulations to the panelists for their excellent sharing of work and experiences on this very interesting and thrilling topic. While there is no denying that AI offers opportunities and targeted solutions to issues of indigenous people across the world, such as the use of machine translation, data carving, or preserving indigenous language and cultural histories. We have the problem that the truth remains that its rapidness is overwhelming and leaves more challenges for indigenous people. Coming from Africa, we have the fear of this risk, particularly in an era of cultural erasure, data extraction, marginalization of our languages, misdiagnosis across spiritual practices, and of course appropriation of traditional knowledge without consent. In recognition of this, we particularly, in the light of the non-recognition of indigenous people in Africa, we are recommending that the EFTA mechanism engage with African government primarily on the call for recognition in Africa and the provision of legal protection for indigenous people in our national laws. To enhance the legal protection of individuals within the context of the AI discourse. Another challenge remains the issue of public awareness. Where other communities are already ahead in terms of being aware, I think that many of us in Africa still lack the adequate information and necessary awareness about AI. And we think that a greater commitment to public awareness is necessary. To drive our inclusion, because without being informed, without that inclusion can be possible. And of course, the next relating to regulatory frameworks. At the moment, African countries are lagging behind in providing necessary laws regulating AI. We recommend that moving forward, there should be integration of EPIC in the development of regulatory frameworks in Africa to protect indigenous within the context of

Thank you for that.

Next speaker, give the floor to Fiji. Thank you, Mr.

Chair. We thank the distinguished guests, panelists for the insightful discussion. Fiji has 4 perspectives, 2 of which are self-explanatory, and if I may read one, the intersection of history, law, and innovation, how they have shaped us. Second, understanding the retrospective echo of historical laws, and as we try to look back and correct the past wrongs, AI is about interpreting the future for the future. The third perspective is on allowing traditional cultures develop the vision for cultural continuity. Technology should serve to support and amplify living cultures rather than replace them. Traditional practices are not static relics of the past; they are evolving, dynamic ways of life that deserve the space to grow on their own. Preventing digital redundancy is a must.

There is a delicate balance between digitization and displacement.

Second, supporting of natural evolution. The goal of integration is to offer supportive technology that communities can adopt at their own pace. On respecting intellectual and creative heritage, generative models should learn to recognize and honor the unique communal nature of indigenous art, language, and ecological knowledge ensuring these expressions are safeguarded and attributed respectfully. The fourth is on the constructive framework, which requires a general affirmation— the general affirmative steps towards equity, which includes the co-design partnerships, supported technical capacity, and sovereign first data students. I thank you.

Thank you very much for that. Next speaker, Consejo Indio Exterior. You have the floor.

Thank you, Madam Chair. History was written and taught by the invader It's their history and therefore it's false and inconclusive. That's why it's necessary to work from a different assumption, the word of Indians. We are a fundamental spirit. Our bodies, our brains, our emotions, everything changes. Everything disappears when we die, but our spirit stays the same. And if you work on that spiritual basis, it doesn't matter if you're struggling There's something you can build on because our spirituality is a fundamental part of what we are. It's our very identity. The English colonialists displaced from the lands and exterminated the Abiyagala Indians. Their doctrine based on America for the Americans was a clear example of the genocide of the When used in the kindred people in Palestine, our lands don't belong to them. Our past doesn't belong to them. Our destiny is not theirs. Our resistance is ours. Donald Trump, abducted by the Zionist Israeli agenda, who he owes his loyalty and his and who had find and owes them for financing his election campaign with his policy of. Imperialism cuts funds for educating indigenous people, fosters disappearance of languages to promote only the English language, takes funds pretending to— and splits our hearts in two and prevents the construction of water and sending money to Wall Street and sending agents against immigrants, against our indigenous peoples, but our voices shall not be silenced. That's why it is our responsibility to fight for freedom so that Donald— to make Donald Trump persona non grata in all territories and to make him go back where he came from. Thank you.

I give the floor to Paulo Resource Institute.

Ali, malungil tu tao. Good morning and thank you, Chair. I am an OHCHR Indigenous Fellow from the Republic of Palau. As a large ocean state in the Pacific, one of the greatest threats to our marine heritage today is the extraction of our marine genetic resources. Through environmental DNA and artificial intelligence, corporations and researchers no longer need to physically harvest our reefs. They simply sample water to extract digital sequence information. Because marine life drifts across ocean currents, our genetic heritage is easily digitized in international waters and appropriated without our consent. This creates a profound injustice. Those with the capital to build AI platforms become the gatekeepers of biological discovery, while Indigenous peoples who have stewarded this biodiversity for millennia risk being excluded completely from the process. Allowing our genetic heritage to be absorbed into global AI datasets without our consent or equitable benefit sharing deprives us of our cultural heritage, economic opportunities, and self-determination, which violates Articles 3, 29, 31, and 32 of UNDRIP. As Indigenous peoples, we cannot welcome a future of digital feudalism, where control falls into the hands of whoever has the most powerful algorithms. Therefore, we call on EMRIP to develop targeted expert guidance on applying free, prior, and informed consent to the digitization of marine genetic resources and to advise member states to explicitly center Indigenous data sovereignty in the implementation of a new High Seas Treaty and similar global digital sequence information frameworks. Thank you.

Thank you. Next speaker, Botanical Society of South Africa.

Thank you, distinguished Madam Chair, panel, and Indigenous elders and leaders from across the globe. My name is Saba Zahara Honeybush. I'm from Cape Town, South Africa, born within close proximity to the historic Hoerikwaggo Mountain, which is also known as Table Mountain and known as the mountain within the sea. We are witnessing a new wave of extraction of land, water, data, knowledge, and even space. Indigenous peoples are facing these same patterns through artificial and superintelligence, digital infrastructure, surveillance systems, and emerging technologies. Technology may replicate images, voices, and behaviors, but it cannot replace the human soul, spirit, cultural connection, and lived knowledge of Indigenous peoples. Today, much of the world's data from the global, Global South is stored, processed, and governed in the Global North, raising urgent questions of digital inequality, sovereignty, accountability, and justice. Data cannot continue to be extracted, from our communities and controlled elsewhere without our participation, ownership, and consent. Indigenous peoples must have control where our data is stored, who accesses it, how it is used, and under which laws it operates. Digital identity systems must not become tools of surveillance, cyber exploitation, or exclusion. The solution is not only access to technology, but Indigenous-led digital systems, community networks, and locally governed infrastructure that allows communities to become owners, creators, and decision makers in their digital future.

Thank you for that statement. Unfortunately, we're forced to interrupt you because all Participants should be able to speak today. Next speaker is Association of Reintegration of Crimea. Next speaker, Indian Social Institute.

Respected Chair, distinguished delegates, and indigenous representatives, Johar. I, Juhi Priyanka Horo from Munda indigenous community, am honored to speak on behalf of India's peoples from India. Artificial intelligence is rapidly transforming the way knowledge is created, shared, and governed. While AI presents opportunities, it also raises profound concerns for Indigenous peoples. Our knowledge systems, oral traditions, cultural expressions, and languages, and sacred practices are not merely data. They are living heritage, collectively owned and deeply connected to our identities, territories, and responsibilities to future generations. AI systems must not extract, reproduce, or commercialize Indigenous knowledge without our free, prior, and informed consent. India is home to over 700 groups with immense indigenous diversity, constitutionally and categorically recognized as scheduled tribes under Article 342, safeguarding unique languages, ecological knowledge, and cultural traditions. AI can become a powerful tool to preserve endangered languages, document traditional knowledge under community control, strengthen indigenous education, and support climate resilience. However, these benefits will only be realized if indigenous peoples of India are recognized as equal partners, not merely subjects of technological innovations. The future of AI must not be built at the expense of indigenous peoples. We call upon member states, UN agencies, technology companies, and research institutions to ensure meaningful indigenous participation in AI policymaking. safeguards to protect Indigenous data sovereignty and intellectual property, invest in Indigenous-led infrastructure and capacity building, and promote ethical AI that respects cultural values, human rights, and biodiversity. Thank you, Chair.

Thank you for that statement. Next speaker, give the floor to Venezuela. Thank you very much, Madam Chair.

The Bolivarian Republic of Venezuela welcomes this opportunity to participate in this discussion on the impact of AI and new technologies on the lives of Indigenous peoples. For our country, this analysis is rooted in our Constitution, which defines us as a multiethnic and multicultural state. And in Article 124 protects the collective intellectual property of Indigenous knowledge. It expressly prohibits any patent or undue appropriation of that knowledge. We recognize that these technologies provide valuable tools for the preservation of languages, for drafting public policies, and for mapping of territories. However, they can entail risks that we mustn't ignore. Technology and digital data cannot under any circumstance replace or impose or be imposed upon the cosmovision, the life, the way of life that are specific to the communities. If we allow external algorithms to interpret the history of Indigenous peoples, or if we allow their traditional knowledge to be used without their consent for commercial purposes, we would be opening the door to a new form of digital colonization of their knowledge, their identity, and their ways of life. Ancestral knowledge belongs collectively to Indigenous peoples. Therefore, the technology must adapt to them and not the other way around. Therefore, any innovation must be developed fully respecting the free prior informed consent of Indigenous peoples, ensuring that the technology preserves and strengthens their voices without taking ownership of their knowledge or distorting it or replacing it. Venezuela reiterates its commitment to ethical and inclusive technology in the service of human rights and the free determination of our original peoples. Thank you very much, Madam Chair.

Thank you. Next speaker, Khmer Krom Federation.

Thank you, Madam Chair. On behalf of the Khmer Kampuchea Krom Federation, we wish to highlight the significant impact that artificial intelligence may have on the Khmer Krom, the indigenous peoples of the Mekong Delta. Vietnam continues to deny our indigenous status, classifying us exclusively according to state-imposed categories. When artificial intelligence systems are trained on official records, government-controlled information, and prevailing historical narrative, they risk, they risk perpetuating this denial and presenting it as objective fact. Search engines, automatic translation systems, educational platforms, and generative artificial intelligence may represent the Khmer Krom only as an ethnic minority, omit our indigenous identity, or exclude the name of our ancestral, uh, homeland, Kampuchea Krom. This is not merely a technical oversight It constitutes a new form of digital assimilation. Artificial intelligence also presents significant risk to human rights defenders. Vietnamese authorities currently monitor peaceful online expression, search mobile devices, scrutinize social media activity. and use digital information against activists and Buddhist monks. Mr. Thay Chantra and Mrs. Chao Thi Tien, a 72-year-old Khmer Krom woman, were arrested in April and June this year for peacefully exercising their indigenous rights.

Thank you for that statement. Next speaker, Commission National— Joseph Marango.

Commission Nationale des Droits de l'Homme.

Thank you. I want to thank the chair. I'm Joseph Marongo, First Vice Chair of the National Human Rights Commission of Congo. The National Human Rights Commission believes that relations between indigenous peoples and AI remains troubling. Languages, knowledges of Indigenous peoples have often been threatened, as is their access to basic services like health, education has often been difficult or almost made impossible due to the lack of appropriate infrastructure. So AI could be seen both as a big opportunity for the population, but also as a threat. AI could help preserve indigenous languages, protect traditional knowledge, sovereignty over territory, and access to fundamental services. But this undeniable contribution of AI is not without risks. The risks linked to AI could pose a threat to Indigenous peoples, a threat to knowledge, use of data, violations of privacy, and it could heighten inequalities. To guarantee the rights of Indigenous peoples in the face of AI, 4 principles emerge at international level, including FPIC, Data sovereignty, participation in governance, and benefit sharing. Besides these risks, more generally speaking, there is a need to clarify that for indigenous populations in Congo-Brazzaville, AI could be particularly useful in participative mapping of forest territories. In surveillance or monitoring deforestation in the Congo Basin, in documenting traditional languages, and in improving access to education and public services. And that brings me to the end. Thank you.

Thank you. Next speaker is Finland.

Madam Chair, distinguished members of the expert mechanism. Indigenous representatives, and dear participants. Finland thanks the Expert Mechanism on holding this important discussion on artificial intelligence. As we all know, AI holds both risks and opportunities. AI can be a tool to support the revitalization of Indigenous languages and the preservation of Indigenous cultural heritage, and can help access educational resources and strengthen the transmission of knowledge to future generations. As an example of the opportunities AI provides, the Finnish broadcasting company Yle already uses AI tools to support the preservation and accessibility of the North Sami language. Additionally, many Finnish universities are developing language models to help preserve the Sami languages. With these opportunities in mind, we must also note the various risks that AI possesses. Algorithm bias can reproduce harmful stereotypes. The lack of meaningful Indigenous participation can lead to exclusion of Indigenous knowledge and perspectives. And a lack of data control can lead to the use of Indigenous knowledge without the consent of Indigenous peoples. AI development must be human rights-based, and Indigenous peoples must be represented in the process to protect both their rights and interests. Distinguished panelists, how can states best protect the rights of Indigenous peoples in the development, deployment, and use of AI tools? Thank you.

Thank you.

Next speaker, Wai Papa Rau, University of Auckland.

Tēnā koutou katoa. Ko Michael Steele me tōku ingoa, nō ngāti Whātua hau. I'd like to offer my thanks to the panellists for their views and offer some congratulations again to the new Chair. Tēnā koe te rangatira. Our speaker is Tangata Whenua from Aotearoa and is a participant of the UN Voluntary Fund for Indigenous Peoples.

I want to thank the Secretariat for their ongoing support of this programme.

I'll draw upon the work of Te Kahui Raronga and Mana Raronga in Aotearoa, who have been advancing Māori approaches to data governance and artificial intelligence. I acknowledge the contributors in these groups wholeheartedly as enabling appropriate leadership and guidance in Aotearoa for this topic. The experience of Māori in Aotearoa highlights an important lesson: responsible AI is about governance, access, reciprocity, authority, and self-determination. Te Kāhui Raraunga Māori data governance model treats Indigenous data as a taonga, or treasure, and it places Indigenous leadership practices and accountability at the centre of data and AI governance. This is one practical example of how rights affirmed in UNDRIP can be applied in the context of emerging technologies. Accordingly, I respectfully recommend that the mechanism considers First, affirming Indigenous data sovereignty and Indigenous-led AI governance as practical expressions of self-determination and participation in decision-making. Second, encouraging states to support Indigenous-led AI governance frameworks and Indigenous participation throughout the design, development, and deployment of AI. Third, investing in Indigenous controlled AI initiatives that support language revitalization, cultural integrity, knowledge transmission, and stewardship of lands and resources.

Namahenuhi.

Thank you. Next speaker, Maribel Villeda Cruz. Not present. Next speaker is Colombia.

Gracias, señora presidenta. Thank you very much, Madam Chair. Colombia welcomes the inclusion of this important matter on the agenda of the mechanism. We believe that artificial intelligence can be a transformative tool to reduce historic divides and to promote the development of Indigenous peoples. We note particular opportunities to revitalize Indigenous languages, the preservation preservation of traditional knowledge, territorial and cultural management, and the expansion of access to public services such as health and education. However, these benefits can only fully come to pass if indigenous peoples actively participate in defining the goals, uses, and limits of these technologies. Colombia believes it fundamental that any and all initiatives incorporate respect for free, prior, and informed consent. We also acknowledge significant risks associated with digital colonialism, the undue appropriation of traditional knowledge, and bias derived from low representation of Indigenous peoples in the datasets used to train AI systems. We therefore support the building of multi-level governance frameworks that are based on human rights that involve states, Indigenous Peoples, academia, the private sector, and international organizations, guaranteeing transparency, accountability, and effective protection of Indigenous cultural heritage. Thank you very much.

Thank you for that statement. speaker, Elçin Basol, organization Izmir Circassian Cultural Association.

Thank you, Madam Chair. While speaking about AI technologies and their impact on peoples, I would also like to address the political and biased data on which AI is often trained. Datasets containing political biases frequently reinforce existing unequal narratives. This problem is particularly evident in the case of my homeland, the Republic of Abkhazia. For more, for more than 15 years, Georgia has refused to accept the Republic of Abkhazia as an equal party within the framework of the Geneva International discussions. The exclusion of one of the parties directly affected by the conflict weakens both meaningful dialogue and the prospect for peace. The political reality is this: if Georgia genuinely seeks a solution, it must first recognize the APAZ as a party and engage with them at the negotiating table. At the same time, Georgia's official discourse, its practices that disregard human rights, and frequent hate inciting publications in the Georgian media shapes the one-sided datasets on which AI is trained. The consequences of this situation are not merely political. The non-recognition of documents issued in Abkhazia restricts the rights to education, work, and freedom of movement. These are concrete human rights problems faced every day by the Abkhaz people. AI systems trained on incomplete or politically selective data risk normalizing these inequalities rather than eliminating them. While AI is increasingly used in border security or even warfare, it is unacceptable for an entire people to be ignored because of document recognition and approval procedures. Thank you.

Thank you very much for that statement.

Next speaker.

Next speaker.

Acehn Sumatra Indigenous People.

Right here.

Madam Chair, thank you so much for the time. My name is Tengku Fajri Kurung, representing the Aceh Sumatra indigenous people. AI has become an important tool for ordinary indigenous people all around the world, especially in Aceh. Many young people use AI to create their videos. However, instead of being encouraged to use technology peacefully, many people are afraid. We have been cases where Acehnese people who create AI-generated content or news AI to discuss this issue that affecting Aceh and raising awareness of issue about Aceh. Later, visit at their homes by member of the security forces and intelligence officers. Is this just happened like 3 days ago in my hometown? Is it, uh, it happened to me too. Especially, uh, they, they block all the social media and everything. These actions create fear and discourage indigenous people from using AI and digital technology to express their views. AI should never become a reason for intimidation. It should empower indigenous people to protect our culture and strengthen our freedom of expression. I respectfully call upon the state and the United Nations to ensure that indigenous people can use AI safely, freely, without fear of intimidation. Thank you, Madam Chair.

Thank you for that statement.

Next speaker is Cuba.

Madam Chairperson, the development of AI is a global trend that is irreversible. The many benefits linked to these technological advances have the potential to contribute to transmitting cultural, linguistic, and historic wealth and to the preservation of ancestral knowledge of Indigenous Peoples. It could also be key to accelerating community governance and economic development, which would undoubtedly contribute to full implementation of the United Nations Declaration on the Rights of Indigenous Peoples, which we fully support. However, none of these opportunities can be effective if the current patterns of technological inequality and discrimination in decision-making in some— the hands of a few states if this is maintained. Cuba supports full participation of Indigenous peoples in the development, implementation, and governance of AI. It is only in doing things this way that the transformative power of this technology can promote inclusive growth, social development, and the promotion of innovation. Protecting our planet. Self-determination of Indigenous peoples must be extended to AI. Digitization of the data of these communities must respect the principles of sovereignty, free prior informed consent, their right to participate in decisions affecting them, their privacy, and their rights to IP. This technology should not perpetuate colonial patterns. It is important to reduce to a minimum the impact of these technologies on the environment. The vast demands for energy, water, and natural resources that are required by the maintenance of AI cannot have a harmful impact on indigenous peoples' rights to land, territory, and the resources that they have traditionally used or acquired. Cuba supports the development of AI. with strictly regulated and ethical inclusive principles oriented to social well-being. Thank you.

Thank you.

Next speaker, Gugumana, Organization Association Espoir.

June, Jens Batwa, or Hope for Youth Batwa. Merci, Madame la Présidente. Thank you, Madam Chairperson. The Batwa Indigenous people of Burundi are a particularly marginalized group of Indigenous peoples faced with very many problems, including very limited access to education. This situation perpetuates their social and economic exclusion. However, technological advancements, including artificial intelligence, today offer new aspects to change and transform this situation. This debate on Indigenous peoples and AI, for me and for my organization, the Association Espoir pour les Jeunes Batois, offers us a space for vital reflection: how to draw on technological innovation to ensure that the Batois people have equitable access to education and to promote effective recognition of their rights while respecting their culture and combating discrimination against them. The Batois thank the Government of Burundi for the efforts already undertaken, particularly the focus placed on education of Batois children so as to promote the rights of a community that has long been marginalized. Artificial intelligence can genuinely improve access to education and can bolster the rights of the Batwa by combining technology, culture, and social inclusion so that they can genuinely become full actors in their own development. It can also help our government to design adaptive educational tools which respect the Batwa culture, enabling tailor-made learning even in isolated rural areas. AI can also contribute to preserving and to valuing traditional knowledge while facilitating the inclusion of the Batwa in the digital society. However, it is vital that these technologies—

Thank you. Unfortunately, I have to interrupt you.

Next speaker.

Apologies for pronunciation. Please, you raised your flag. Go ahead. Yes, the person raising their hand, go ahead.

Thank you, distinguished experts. I'm speaking representing the Tsimotsetén Regional Council of Bolivia. My name is Milton Cune Sarabia, and I'm a young indigenous person from the Tsimane people. I'm a student of education. I'm here in this forum to express the deep concern of my people at the advance of AI and the new technologies. While the world is moving forward, in a modern way. And talking about automation, communities like mine lack basic infrastructure. We are suffering from historic discrimination and exclusion. For us, AI represents a double— a twofold risk. It's going to take on our ancestral knowledge and cultural mobilization. We denounce the fact that AI models are feeding on and making profit from indigenous culture and languages without our free prior informed consent. They are violating our rights to intellectual property, which is collective. The technology cannot become a tool for digital colonization.

Thank you for this statement. I'm forced to interrupt you. Next speaker, Samamti.

Thank you, Madam Chairperson. Ladies, gentlemen, delegates, representatives of indigenous peoples, my name is Sema Ptinabon and I'm speaking on behalf of the Watejuma. This is a French organization working to promote the rights and peoples of indigenous Indigenous peoples of the Chittagong Hill Tracts in Bangladesh. The United Nations Declaration on the Rights of Indigenous Peoples recognizes our right to preserve, transmit our languages and knowledge, and our cultural heritage. Today, artificial intelligence is opening up new windows for opportunities, opportunities translating our songs, our stories, and our heritage from our communities is enabling our history and our stories and our culture to be better known. Within our association, these tools are helping us already to translate information and to communicate with institutions to strengthen cultural ties and exchanges. However, indigenous people must be fully involved with the development of this technology. Traditions that a— or statements that AI translations that AI makes must be checked with the originals in order to ensure that the, uh, that they, they respect our, our knowledge and our languages. We encourage universities and companies to support the preservation and the valuing of indigenous languages. In direct partnership with the people concerned, for Indigenous people, AI must be a tool of transmission of dialogue and of recognition. Thank you very much.

Thank you very much. Next speaker, Owens.

Chairperson and distinguished members of the expert mechanism delegates. Tēnā koutou katoa. Ko Tamawera Owens aho. The question is no longer if AI will reshape our societies, but how.

For 476 million Indigenous people, how has never mattered more.

The United States Cloud Act lets US authorities compel any company under American jurisdiction to hand over your information, whether it is stored—

your medical records, your language archives— without your consent, without your knowledge, without your due process.

Process at home. It has a name: digital colonialism.

Free, prior, and informed consent cannot survive a foreign subpoena. As a sovereign intelligence founder, I build on 3 pillars. First, personal data ownership. Your data belongs to you. Your medical records reach only the doctor you choose. Your health records follow you anywhere around the world, secure, complete, and yours. And what is true for a person is true for, for a people. Article 31. An AI model trained on an indigenous language must answer to that language's people.

Consent built into the architecture, not requested after the harvest.

Second, trust into the hardware itself.

AI now lives in a silicon.

If those chips hide backdoors, no software rule can save us. We need international standards for microchip screening, and personal databases belong in the home, secured by multi-factor authentication and cryptographic keys only you hold, not in someone else's data centre. Hardware stands on land. Infrastructure on Indigenous lands requires Indigenous consent, participation, and benefit. Articles 25 and 26. Third, human-centred governance. Algorithms must never have the final word on decisions that belong to the humans— to humans. A court sentence, a diagnosis, a hiring, a military strike. Automation can advise, it cannot absolve us of responsibility. The final call rests with of the accountable human being, the professional who holds the accreditation, always. Bias and surveillance fall hardest on Indigenous communities. Governance must mean, must mean what Articles 18 and 19 require. Indigenous people in this room, nothing about us without us. These are not anti-technology positions, they are pro-sovereignty.

Thank you for that statement. Next speaker. Society of Native Nations.

Jamie Williams.

Jamie Williams.

Thank you, Chairperson, and for the opportunity to speak. My name is Jamie Williams from the Miami Nation of Indiana in the United States. I am speaking on behalf of the Society of Native Nations. The annotated summary provided does not reflect the true impact of artificial intelligence on Indigenous peoples, particularly in the way the technology is powered. These massive-scale projects are rapidly being constructed across the United States on the lands and territories of Indigenous peoples without free prior and informed consent. My peoples came out of the water, the St. Joseph River in southeast of Lake Michigan. Along that river, there are already several data centers. My indigenous tribes from across Turtle Island share a similar story that these sacred waterways and lands are being used for AI data centers. I was once convinced artificial intelligence could help my tribe rebuild our language and culture, but artificial intelligence is not a sustainable solution. There is no currently available technology to power AI without poisoning the water, drying out the rivers, and causing further contamination in the territories of Indigenous peoples that are disproportionately impacted already. In the United States, the data center industry remains completely unregulated. These are unbelievably massive buildings reinforcing our reliance on fossil fuels and pumping toxins into the air, land, and water 24 hours a day. Meanwhile, our indigenous peoples are already harassed and unjustly detained by government agencies infiltrating the cities. It is our recommendation that MRIP include the study and impacts of AI data centers that power artificial intelligence and their direct effect on indigenous peoples' land, water, and ways of being. We further request that MRIP provide technical advice to states on ensuring that AI development—

Thank you for that statement. Next speaker.

Tai Chi Ming Cha, Organization Hmong Heritage.

Tai Chi Ming Cha from Hmong Heritage Organization.

Thank you, Madam Chair and EmRatThich, for giving us this opportunity to speak for the very first time in the United Nations. My name is Tai Chi Ming Cha. the president of Hmong Heritage. Speaking on behalf of the Hmong indigenous people, M-O-N-G, the Hmong are an ancient people with deep historical roots in Northern Asia and Eastern Asia who are now scattered around the world. We possess our own distinct history, culture, and language, yet our Hmong people are misclassified.

Thank you.

for being other racial ethnicities. As an example, when searching for Hmong people in the Asia digital platforms and AI systems, Hmong is either not found or being classified as Mongolian. In the West, search engines and AI systems misrepresent Hmong people for being a subgroup under Hmong-Miao identity and defines our Hmong name into derogatory meanings. Hmong and Hmong-Miao have distinct histories, separate cultural lineages, and different identities. Misrepresentations has caused lasting harms, reinforcing discrimination, exclusion, and dehumanization. AI systems are trained on existing information that often contains inaccurate and biased representation of the Hmong, further entrenching the erasure of the Hmong people. To stop the erasure and safeguard the Hmong language, history, identity, and cultural heritage, urgent official recognition and corrective measures are needed. We respectfully request your support to include the Hmong Indigenous people into the AI initiatives to engage with AI systems globally to make corrections to the false narratives about the Hmong so our language, history, and cultural identity can be preserved for future generations.

Thank you for that statement. Unfortunately, I had to interrupt you. Sorry about that. And before— I'd like to ask colleagues, the members of the EMRIP Before I go to them for comments, I want to give the floor first from the list of speakers to the organization Winsports International.

Respectful greetings. Our delegation would like to raise a few questions for the panel members. What is being done to ensure respect for sacred protocol? And the requirements of such when establishing or accessing a virtual or artificial elder? Many of our sacred teachings call for a strict compliance of ethical and moral conduct. Secondly, with the rise of misinformation and disinformation, how and what safeguards should be developed to protect from the creation of manufactured Falsified consent or AI-generated free, prior, and informed consent? How are or do potential infrastructure investments into data centers measure for compliance with UN and American declarations on the right to development? How are these investments measured for the protection of lands, territories, and resources, including water? How are the investments monitored to ensure ethical investing and respect for human rights under the Business and Human Rights Guidelines and Principles? Lastly, specific Indigenous ethical investments and policies for sustaining development on artificial intelligence and data centers. How should Indigenous-owned corporations participate in technology advancement in their exercise of the right to development. In a world where non-Indigenous tech companies are attempting to own our information, how can we participate, protect, and safeguard our own sacred intellectual property in a way that prioritizes safety of knowledge and ethical economic interests of Indigenous peoples?

Thank you. Thank you for that statement. I turn to my colleagues, the distinguished members of the EMRIP, of the mechanism. Would you like to comment? Madam Chair of the EMRIP, Val Mayne, go ahead.

Thank you, Madam Chairperson. The long reach of artificial intelligence has placed pressure and challenges for Indigenous peoples and Indigenous rights. Although there may be opportunities and benefits, of AI for Indigenous peoples, AI is perceived as a new coloniser, therefore emphasising the need for adequate recognition and protection of Indigenous peoples' rights when, when intersecting with AI. It is evident that AI is accelerating the rate at which society consumes and disseminates information irrespective of its accuracy. This unprecedented pace, particularly in relation to the circulation of AI-generated generated interpretations of Indigenous knowledges highlights the critical importance of Indigenous perspectives and dialogue. Articles of the UN Declaration provide clear support for Indigenous peoples' rights when intersecting with AI, which we have heard from the floor today. However, as emerging technologies such as AI continue to transform how knowledge is collected, used, commercialized, and disseminated, it is It is vital that we ensure Indigenous peoples' rights continue to be meaningfully upheld and protected in this new age. On that note, I welcome the comments from Finland on seeking advice from Indigenous peoples, and in this light, support the Māori data governance models also mentioned from the floor as a key example that can achieve this. Thank you, Madam Chair.

Thank you very much. Uh, Anex Cunningham, please. Thank you very much, Madam Chair. I'd like to thank the panelists for a very timely, enriching debate. That discussion makes it clear that AI cannot be developed in the margins of human rights of Indigenous peoples. AI offers great potential, but to strengthen indigenous languages, preserve traditional knowledge, and support transmission of our cultural heritage. But also, it could worsen historic inequalities when data, knowledge, and indigenous identities are used without our consent, and when systems reproduce bias and exclusion. I'm also grateful to the speakers in the chamber stating their concerns and stances. The INRIP believes that governance of AI is fully aligned with the UNDRIP. This means guaranteeing FPIC. Recognizing Indigenous data sovereignty and ensuring full effective participation of Indigenous peoples in design, development, and oversight of these technologies. I'd like to make— leave a thought for the panelists, a question: how can we move forward to general principles towards concrete accountability mechanisms that ensure that AI respects collective rights of indigenous peoples to the land, natural resources, self-determination including forms of governance, political governance among others, which can yield the benefits of those technologies? How can we ensure that the benefits of those technologies can be shared fairly with indigenous communities? Thank you.

Thank you very much, Madam Cunningham. I must confess I don't see any other— from the screen, I don't see anyone else of my colleagues wanting to take the floor. Is there anyone else wanting to take the floor among the EMRIP members? Thank you very much. Then we can return to the distinguished panelists. Give them a chance to take two minutes to make concluding remarks and answer some of the questions that were put by the distinguished members of the mechanism. And first, I give the floor to Madame Storslet. Please go ahead.

Thank you, Chair. I would like to thank my fellow panelists and those. Taking the floor in this important discussion. Today's interventions have clearly highlighted both the profound opportunities and the serious risks that AI presents for Indigenous Peoples. Indigenous Peoples must be meaningfully involved in all stages of AI development and governance. Our rights must be fully respected across the entire lifecycle of AI. This requires that free, prior, and informed consent be upheld not only in relation to land and natural resources, but also in relation to Indigenous peoples' data, languages, knowledge systems, and within technological development. Indigenous peoples must be meaningfully included in the development, governance, and regulation of AI, and develop our own system of data governance and AI governance. Governance grounded on our values and protocols. We cannot remain on the margins while decisions are made that directly affect our peoples and societies. Our perspectives, knowledge systems, and priorities must be reflected in both national and international framework. As said before, nothing about us without us. I also share the concerns expressed regarding the impact this development has on indigenous peoples' lands, water, and natural resources, and how this is affecting indigenous communities and our well-being. These matters must be addressed, and I encourage EMRIP, member states, and technology corporations to include this in their future work on this issue. We must ensure that AI development does not come at the cost of Indigenous peoples' rights. We must ensure that AI becomes a tool for empowerment, not a source for further marginalization or digital colonization. This means actively developing policies and strategies that enable Indigenous peoples to benefit from technological while safeguarding our rights, knowledge, cultures, and way of life. It is important to acknowledge that AI holds significant opportunities for indigenous peoples within language, transmission of knowledge, innovation, economy, and social services. However, these benefits can only be realized if indigenous peoples are included as equal equal partners in shaping and governing these technologies. This requires structural and sustainable support of Indigenous-led institutions and experts, including both technological developers, our language and knowledge holders, and capacity building within Indigenous communities. From a Sami perspective, we stand ready to contribute to this work through international cooperation within the UN, through our institutions, and through partnerships with other Indigenous Peoples. The path forward must be rooted in human rights, guided by the UNDRIP, and shared through genuine partnership. Olu Gito.

Thank you very much. And for concluding remarks, I'd give also the floor to Emily Pugin to make her Thank you, Madam Chair.

Colleagues, as we consider the future of artificial intelligence, we must ensure that technological progress strengthens rather than diminishes the rights, cultures, and knowledge systems of Indigenous peoples. AI governance must be grounded in partnership, trust, and accountability. It must recognise Indigenous peoples not only as people that are affected by technology, but as leaders, as right holders and decision makers in shaping how technology is designed, used, and governed. This requires respect for self-determination, free, prior, and informed consent, Indigenous data sovereignty, and the protection of cultural heritage, language, and traditional knowledge. It also requires practical action to close digital divides and support Indigenous-led innovation. Australia welcomes continued dialogue through the expert mechanism and with Indigenous peoples, member states, and partners. Together, we can help ensure that AI develops ways that are safe, inclusive, culturally respectful, and firmly anchored in human rights. Thank you, Madam Chair.

Thank you very much for that statement. Now for concluding remarks, I give the floor to Vasily Niemitschkin.

Thank you, Madam Chair. In conclusion, I'd like to thank once more the Emirate for their invitation. Presentation and thank all other panelists and speakers from our sociocultural region. Many of the statements strengthen our belief that developing AI should take place based on ethical principles, upholding the rights of Indigenous peoples when it comes to collection and use of traditional knowledge, uh, the depicting depictions of languages of Indigenous peoples. For example, from our socio-cultural region, the statements that show that AI can help protect development of Indigenous languages, including niche languages, and some of the examples given— written translation, for example, word recognition, linguistic modeling— they all allow us to create a linguistic repository of Indigenous languages. Traditional knowledge of Indigenous peoples is a great asset which remains to a significant degree insufficiently documented, and AI could help preserve, systematize, and transmit that knowledge, which obviously requires strict ethical regulation. In conclusion, to respond to the question from Madame Cunningham on measures, in my opinion, subjectively, the time has come to put together an international document takes into account the rights of Indigenous peoples when using AI technologies. Such practices, such experience already exists when developing the Global Plan of Action on protecting Indigenous languages with the participation consultations with all socio-cultural regions in the world. And in my opinion, I think the time has come for UN Indigenous mechanisms to initiate such a document. So issues of AI and taking into account the rights of Indigenous people in that regard.

Thank you, Mr. Nemetschkin. And for concluding remarks, I give the floor to, to, to Mr. Ariston.

Madam Chair and distinguished members of this panel, thank you to the EMRIP Committee for inviting me to I think I'm just responding to what I heard today and learned from everyone speaking, that uppermost in most people's comments is the extraction and use of Indigenous knowledges without Indigenous peoples shaping the process and their free prior and informed consent. AI not only proposes a new form of colonialism accelerating the circulation of information, as you just said, whether or not it is accurate or respectful, which is why this is a very important discussion to have. Thanks to my cousins from across the great ocean for their words in saying AI development must commit to protecting ancestral intelligence, not come at the expense of our knowledge systems or the people entrusted with protecting knowledges and passing it on in our actual living communities. Cultural continuity is important, and these technologies serve, support, and amplify living cultures rather than replace them. Here's an example. Building large language models in my community, which is unrecognized by the United States, without the input of the majority of the people in my community who were themselves prevented from speaking or knowing language and culture due to colonial laws, places the actual community in a strange backseat position to the AI who will now have all this knowledge. But the AI is only synthesizing material it can get its hands on, and most of that material, even if it may be in Hawaiian or if it's in translation, may still be produced by people who are not from my community, who are not indigenous. So what is it feeding us, and what is it homogenizing in that process? These are questions that are absolutely lost in these conversations where nuance and depth is required. Another observation, when we discuss ethical frameworks, these are often not based out of our languages and traditional knowledge systems. There are no one-size-fits-all ethical frameworks, so I would like to see discussions about the ethics. I think we need to create collaborative spaces for us Indigenous peoples to work together on these issues and protect our knowledges, languages, and homelands to mitigate the risk of digital colonialism. And to promote Indigenous legal protections, further recognitions of communities that strengthen our sovereignties, how our knowledges are conducted in AI. We need a standing working group on this issue to conduct studies of impact of AI data centers in Indigenous communities. And I agree with the last speaker, an international global document that's created to protect Indigenous rights to our knowledges in the age of AI.

Thank you. Thank you, Madame Arista. I give the floor for final comments. I give to Mr. Olivier Allais.

Madame Chairperson, by way of conclusion, allow me just to link various threads from this morning's discussion. Indigenous peoples are not late arrivals the digital era that need to be caught up. They are the holders of rights of their own systems of knowledge, their languages, and their governance traditions that technology should take into account and not the opposite. For the ITU, and in line with the provisions governing the nature of members of the ITU, that is done through 3 commitments that I wanted to note. We continue to support connectivity by Indigenous Peoples because no right mentioned here can come to pass without that. We will support Indigenous control over data and linguistic resources that feed AI systems by basing our work on models such as community licenses, which maintain governance where it is. And we encourage Indigenous experts to participate in our standards drafting processes so that they can contribute to shaping ICT norms and standards and AI from first stages, first steps, rather than just simply commenting on them post facto. Also, to answer a question from the room, the C2PA Coalition for Authenticity in the ITU are collaborating actively on the origin of multimedia content to combat disinformation and deepfakes. Fakes. This is structural work which will not succeed— or it shall only succeed, rather— if Indigenous peoples are present, equipped with resources and understanding, in the global dialogue on AI governance, in the AI Summit for the Common Good, and in each and every one of the standardization meetings to come in the future at the ITU. I want to thank the Emirate for having given me the opportunity to listen and to be held accountable. The ITU welcomes the ongoing conversation over and above what's happened today. Thank you.

Thank you very much. Our next speaker for concluding remarks, Faidou Doumarpo. You have the floor.

Thank you, Madam Chair. Again, greetings of peace and assalamualaikum. As GANRI's closing statement, we encourage further reflection on ethical and legal frameworks that go beyond mitigation of harm and towards positive recognition of Indigenous knowledge systems as contributors to AI governance itself. Indigenous epistemologies, governance models, and relational understanding of data and environment can offer important alternatives to dominant technological paradigms. AI is not neutral. It reflects the values, power structures, and choices of those who design and govern it. Without deliberate action, there is a real risk that AI will deepen existing inequalities faced by Indigenous Peoples. With meaningful participation, empowerment, accountability, non-discrimination, and legality, the rights of Indigenous People can be further promoted and protected. GANRI also continues to strengthen its internal engagement on these issues, including through its Caucus on Indigenous Peoples established in 2004, which supports coordinated NHRI action and peer learning in this field. We therefore encourage MREP, states, national human rights institutions, Indigenous Peoples, the private sector, and UN partners to work together to ensure that AI development is firmly grounded on the UN Declaration on the Rights of Indigenous Peoples and international human rights law. Thank you. Thank you very much.

I must note that the distinguished panelists were very disciplined, sticking to the time limit. That— and that for that reason, we have a bit of time to be able to take 2 or 3 more statements from the chamber, from those representatives of Indigenous organizations that didn't have to speak during the session. So let's just take advantage of this. And I'll give the floor to Action for Human Rights.

And if you— thank you, Chair, distinguished MREP experts, honored delegates, and colleagues. I speak as we confront a new frontier of dispossession: digital colonialism and its twin, unaccountable AI. Indigenous data, cultural knowledge, languages, land stewardship records, health and biometric markers are being harvested, commodified, and encoded into systems that neither reflect our values nor respect our collective rights. Left unchecked, algorithmic systems will reproduce and amplify historical injustices. We must move beyond mitigation to structural redesign. For this, we propose following actionable measures in this forum. First, statutory Indigenous Indigenous Data Sovereignty Council with veto on contracting authority over public sector AI projects and government control over privately owned AI projects focusing on Indigenous people. This council must be resourced, technically supported, and empowered to enforce free, prior, and informed consent. Second, mandatory Indigenous impact assessments for any AI system touching Indigenous lives or lands. These must be binding in procurement, published with mitigation plans, and include enforceable remedies and timelines. Third, community-owned federated data infrastructures and sovereign cloud options that keep data physically and legally within Indigenous control. Funded seed projects should demonstrate how local ontologies and governance rules can be encoded into interoperable systems. Fourth, legal recognition of collective data rights and benefit-sharing mechanisms. From data trust to contractual royalties, so that value derived from indigenous data returns to communities and supports self-determination. If we center indigenous authority now, AI can become a tool for cultural resilience rather than a new instrument of dispossession. I thank the expert mechanism for this opportunity.

Thank you very much for that statement. Next speaker.

Organización Maya Wayob Cambio, Mr. Brian Guillermo Batzin.

Distinguished Chair, members of the Emirate, participants, our organization thanks the mechanism for giving us an opportunity to participate in negotiating and dialogue panels. I want to draw attention to growing relationship between between AI and Indigenous peoples' rights. Although this technology creates opportunities for education, revitalizing Indigenous languages and access to knowledge, it also poses major risks including undue ownership of traditional knowledge, use of data and images without consent, and reproduction of historic forms of discrimination. In line with Article Articles 6 and 7 of ILO Convention 169 and Articles 19 and 31 of the UNDRIP. It's crucial to recognize that indigenous peoples, we have a right to be consulted, a right to participate in decisions that affect us and to protect and control our knowledge, data, and cultural expressions. Therefore, I urge indigenous peoples, especially indigenous youth, to participate fully and effectively in the design, development, and governance of AI. Technological innovation should not become a new form of epistemic extractivism or digital exclusion. I urge this mechanism to continue promoting studies and recommendations foundations that guarantee artificial intelligence based on human rights, cultural diversity, and self-determination of indigenous peoples.

Thank you.

Thank you. Next speaker.

Mr. Olner-Ortiz. Merci.

Thank you. AI doesn't involve— evolve in a normative vacuum. The rights in the UNDRIP fully applied to the digital environment and the norms necessary to govern that technology already exist. Their faithful implementation and the risks are tangible. The risk of data reproducing biases and misimpressions in contradiction with the ban on discrimination. Extraction of knowledge, languages, and Indigenous knowledge without consent perpetuates historic patterns of dispossession using a new means, a dynamic that the International Working Group for Indigenous Affairs has described as digital colonialism. Generational systems that reproduce sacred images are a violation of Article 31 of the Declaration, and surveillance made possible by AR targets increasingly Indigenous defenders, as stressed by the Special Rapporteur in his report in October 2025 to the GA. And yet this technology can carry promise under the control of communities and with their free prior informed consent. They could support revitalization of threatened languages and demarcation of territories and lands. What depends is control. 3 points deserve the attention of the experts. First, FPIC under Articles 19 and 31 governing all use of data, knowledge, and cultural indigenous expressions. Secondly, the Cairo Principles for Indigenous Data Governance promoted by the Global Indigenous Data Alliance should be adopted parallel to any open data policy so that that openness should be sought never to the detriment of indigenous people's ownership over their own data. Thirdly, Indigenous peoples must be involved right from the start and not just consulted once the system has been built. As the OHCHR has said, any process must be led by Indigenous peoples and aligned with the— it should be aligned with the postulate of nothing for us without us. Thank you. Our final speaker, New South. New South Wales Aboriginal Council.

Thank you, Madam Chair. My name is Anne Dennis. I'm a councillor with the New South Wales Aboriginal Land Council, and we stand and support our sisters in the Kimberleys with their statement. Artificial intelligence is already reshaping our societies. The question is not whether AI will shape our future, but whether Indigenous people will help shape AI, or will we continue to shape it without, without our participation? The rights contained in the UNDRIP already apply to artificial intelligence, and the challenge is implementation. This morning, Australia referred to the Australian Artificial Intelligence Institute. Having heard this for the first time, we sought to better understand its work. Based on publicly available information, we have been unable to identify Indigenous researchers or Indigenous-focused projects. If this work exists, it should be visible. It should not— it reflects the continuing gap, Australia's international commitments and domestic implementation. We welcome the government, the Australian government statement The examples demonstrated in Australia's international commitments are not translated into domestic implementation. Close the Gap is failing our Indigenous people. It does not build self-determination, and the project funding provided is not sustainable. And despite Australia endorsing the Declaration in 2009, the standards have never been incorporated into domestic law and continue to be treated largely as guiding principles. To realize these opportunities, Indigenous people must not simply be consulted after AI systems have been designed. We must participate in their design, governance, and oversight from the outset. Always was, always will be Aboriginal land.

Thank you very much for that statement. And I think we already have really gone into depth during this panel and we've heard a lot from those organizations who were unable to speak yesterday. For my part, I want to thank you all for this discussion, state representatives, indigenous organizations, the distinguished panelists today. Repeatedly, we've heard the thought that this is a very timely discussion. And let's hope that we really will be— have drawn attention to this issue and not later than required by the rapid development of AI. We also said that this theme was very complex, maybe endlessly complex, affecting the worldviews of indigenous peoples, traditional values of indigenous peoples, Also, the issue of equal access to technologies, states and indigenous peoples. Then there's the environmental impact of the development of AI in the territories and lands of indigenous peoples. There's also the opportunities that AI offers if properly used. Helping develop, preserve languages of indigenous peoples, protecting cultural heritage, and also many other issues including preserving traditional use of land and resources. I should remind you all that starting 3:00 PM until 4:30 PM in this chamber, there will be a closed-door session of the mechanism and all participants of the session. We'll see you here in— at 4:30 for the next agenda item for the session. I thank you all.