AI and Indigenous Peoples' Communication Rights: Advancing Inclusion or Creating New Barriers?
This paper was presented at the Silicon Valley Peace Conference 2026 in Palo Alto, California, United States, held September 17–20, 2026, under the theme “AI’s Role in Community Development and Peacebuilding in Developing Nations,” and is published as part of the Proceedings of the Silicon Valley Peace Conference 2026.
Abstract
The number of indigenous peoples (IP) is estimated at 476 million living in 90 countries across the world. They speak an estimated 7,000 languages and represent more than 5,000 distinct cultures. Although they comprise only about six percent of the world’s population, IP account for 18.2 percent of the world's extremely poor (World Bank, 2025). Indigenous Peoples hold diverse cultural values, traditions, and historical heritage deeply rooted in their ancestral domains. Despite this richness, they remain among the most marginalized sectors, frequently excluded from socioeconomic and political opportunities. They are also disproportionately vulnerable to environmental, health, educational, and economic crises. In an increasingly mediated world, many IP communities continue to experience digital exclusion. This marginalization has prompted the international community to recognize distinct rights and freedoms for Indigenous Peoples, as articulated in multilateral declarations and agreements. Many countries have also adopted their own laws promoting and protecting rights of IPs. Artificial Intelligence presents both opportunities and challenges for IPs. It offers potential pathways for the exercise and fulfillment of their rights, and for the meaningful inclusion of IPs in sustainable development processes. This paper used qualitative document analysis, framed by a rights-based critical policy analysis, to examine not only the impact of AI on the IP community but more important, whether IPs exercise agency, voice, rights, cultural integrity, data and knowledge sovereignty, and meaningful participation throughout the AI lifecycle. The analysis draws on purposively selected international normative instruments, United Nations and other intergovernmental policy documents, scholarly literature, proceedings of relevant forums, church teachings, documented lessons from development programs and ICT initiatives. These sources are selected based on their authority, relevance, credibility, recency, geographic and cultural diversity, and substantive contribution to understanding the intersection of Indigenous Peoples' rights, artificial intelligence, communication Findings and analyses are also complemented by reflexive practice-based insights derived from the authors' documented experiences in managing ICT initiatives with Indigenous communities in the Philippines and Myanmar. The documents are analyzed thematically through an integrated framework encompassing Indigenous Peoples' rights, human rights, communication rights, cultural rights and sensitivity, AI ethics and governance, data and knowledge sovereignty, and Free, Prior and Informed Consent. It concludes by offering concrete policy and program recommendations for AI engagement of IP communities across the AI lifecycle, toward the common good and with a preferential option for Indigenous Peoples.
Keywords: Indigenous Peoples, Artificial Intelligence, Communication Rights, Human Rights, Cultural Sensitivity, Sustainable Development, Free, Prior, and Informed Consent (FPIC)
Introduction
The number of indigenous peoples (IP) is estimated at 476 million living in 90 countries across the world. They speak an estimated 7,000 languages and represent more than 5,000 distinct cultures (UN, 2026).
Although they comprise only about six percent of the world’s population, IP account for 18.2 percent of the world's extremely poor (World Bank, 2025).
The advent of artificial intelligence (AI) offers many benefits and opportunities to IP. But like any other technology or innovation, AI can also bring issues and challenges. This paper adopts UNESCO's working approach to defining AI as computational systems that process data and information in ways that resemble intelligent behavior and typically include
aspects of reasoning, learning, perception, prediction, planning, or control (UNESCO, 2021).
The UNESCO Recommendation on the Ethics of Artificial Intelligence (UNESCO, 2021, p. 32) calls upon Member States to incorporate AI systems in the "preservation, enrichment, understanding, promotion, management and accessibility of tangible, documentary and intangible cultural heritage, including endangered languages as well as indigenous languages." However, the Recommendation also warns of the "potential negative implications" of AI for cultural diversity and the pluralism of languages, media, cultural expressions, participation, and equality (UNESCO, 2021, p. 35). In addition, the Recommendation recognizes that AI can reinforce and exacerbate existing forms of discrimination, prejudice, and stereotyping (UNESCO, 2021, p. 6).
These dangers are reinforced in the Preliminary Report of the Independent International Scientific Panel on AI: Evidence-based Assessment of Opportunities, Risks and Impacts of Artificial Intelligence (Independent International Scientific Panel on AI, 2026). Among the main takeaways of the report is that “AI systems optimized for engagement and scale are not neutral: they embed primarily English-speaking and global North cultural assumptions that can actively marginalize most of the world’s population” (Independent International Scientific Panel on AI, 2026, p. 40).
This UN Preliminary Report also calls for more inclusive AI development across the entire AI life cycle, including addressing structural imbalances in who develops, defines, owns, and governs AI systems. It also emphasizes the need for investments in AI capacity, infrastructure, and skills across countries and regions, as well as the development of more representative datasets and benchmarks.
Significant principles embodied in the UNESCO Recommendation are echoed in Magnifica Humanitas (Pope Leo XIV, 2026). The encyclical recognizes the greatness of humanity as reflected in its achievements in scholarship, technology, and innovation in advancing human development and promoting the common good. At the same time, it cautions that “each phase of progress has also revealed the ambiguity of tools that can cause harm when not oriented toward the good” (Pope Leo XIV, 2026, para. 4). Magnifica Humanitas thus warns against the dangers of the technocratic paradigm.
Working for a balanced, open, equitable, fair, safe and secure AI ecosystem which highlights the rights of IP requires a “whole of society” approach with active engagement of state actors, indigenous communities, cultural workers including activists, linguists, and technology experts. This multistakeholder approach to fostering a desired AI ecosystem is echoed in the Pact for the Future and its annexes, the Global Digital Compact and the Declaration on Future Generations, which were adopted by consensus on 22 September 2024 at the Summit of the Future (United Nations, 2024a). The Summit brought together over 4,000 individuals, including Heads of State and Government, observers, intergovernmental organizations, the United Nations system, civil society, and non-governmental organizations (United Nations, 2024b).
Recalling Indigenous Peoples’ Communication Rights
International declarations recognize the basic human rights of IP. These rights include communication rights that are embedded in universal human rights documents including those which specifically relate to culture and development.
Among the major documents are the UN Declaration of Human Rights (1948), International Covenant on Civil and Political Rights (1966), the UN Declaration on the Rights of Indigenous Peoples or UNDRIP (2007), Universal Declaration on Cultural Diversity (2001), and the Convention on the Protection and Promotion of the Diversity of Cultural Expressions (2005). One of the most recent declarations was adopted during the UNESCO World Conference on Cultural Policies and Sustainable Development – MONDIACULT 2022.
What constitutes communication rights? The UNDHR identifies these fundamental communication rights to include freedom of expression (including press freedom); freedom of information (right to information); right to privacy; freedom of thought, conscience and religion; right to freedom of peaceful assembly and association; the right to education; and right to participate in the cultural life of the community, to enjoy the arts, and to share in scientific advancement and its benefits.
Culture-related documents, especially the UNDRIP, cover many other communication rights of indigenous peoples. These include the following:
Right to be free from any kind of discrimination (UNDRIP Article 2); Right to the dignity and diversity of their cultures, traditions, histories, and aspirations (UNDRIP Article 15 1);
Right to practice and revitalize their cultural traditions and customs (UNDRIP Article 11);
Right to maintain, control, protect and develop their cultural heritage, traditional knowledge and traditional cultural expressions, as well as the manifestations of their sciences, technologies and cultures, including human and genetic resources, seeds, medicines, knowledge of the properties of fauna and flora, oral traditions, literatures, designs, sports and traditional games and visual and performing arts (UNDRIP Article 31 1);
Right to maintain, control, protect and develop their intellectual property over such cultural heritage, traditional knowledge, and traditional cultural expressions.” (UNDRIP Article 31 1);
Right to establish their own media in their own languages and to have access to all forms of non-indigenous media without discrimination (UNDRIP Article 16 1); and
Right to revitalize, use, develop and transmit to future generations their histories, languages, oral traditions, philosophies, writing systems and literatures (UNDRIP Article 13 1).
Leaving No One Behind Online
Access to technology and internet connectivity is a basic requirement to harvest the benefits of AI.
According to the International Telecommunications Union (ITU), about a third of the world’s people (2.6 billion) do not have access to the internet. They live in rural and remote communities, least developed countries (LDCs), and landlocked and small island developing countries. ITU does not have data on the digital divide gripping IP. However, it may be safe to assume that IP are among the least connected considering that they live in unconnected areas described earlier. ITU also estimates that 15 percent of the rural population live in areas with no mobile coverage, with lesser percentage covered only by 2G network (ITU, 2023).
Among the connectivity challenges that IP face are geographic isolation, reliable internet access, affordability, digital and AI literacy), and access to electricity. An ITU paper reveals that the cost of broadband connectivity remains above the affordability target of two percent of monthly gross national income per capita for a number of LDCs set by the Broadband Commission for Sustainable Development. Meanwhile, so-called deprived, depressed, and underprivileged communities (DDUs) remain least attractive to commercial technology companies because there is no assurance of return on investment or profit.
AI and Language Diversity
According to UNESCO, there are over 8,300 languages worldwide. UNESCO warns that the world loses a language every two weeks. Around 40 percent of the global languages are at risk of disappearing. Languages are not merely tools for communication; they are repositories of histories, arts and cultures, heritage, and identities. Indigenous knowledge on health, ecological and sustainable management practices, peacebuilding strategies, etc. are embedded in local languages (UNESCO, 2024, 2025).
AI can help document, translate, and revitalize Indigenous languages. AI can be mobilized to preserve endangered languages and contribute to multilingualism needed in promoting equitable access to information and knowledge, particularly within the public domain and in cyberspace.
Ajani, et.al, (2023) citing Gwerevende and Mthombeni (2023), noted how harnessing digital tools and platforms (including AI-driven technologies) can foster collaborations, share valuable resources, and revitalize languages in unprecedented ways.
Mobilizing AI for language and content development is challenged by the underrepresentation—or even absence—of most of the world's languages in cyberspace. According to the Observatory of Linguistic and Cultural Diversity in the Internet (OBDILCI), the number of languages represented in cyberspace is estimated to be 343 languages, accounting for only about 4–6 percent of the world's languages (OBDILCI, 2025). Moreover, the ten most represented languages account for approximately 84 percent of online content (OBDILCI, 2025).
These disparities underscore the challenges of ensuring that AI systems are trained on linguistically and culturally diverse data.
The Preliminary Report of the Independent International Scientific Panel on AI (2026, p. 40), citing several research studies, laments that AI models trained on dozens of languages perform well for only a small subset of them. It further notes that recent studies indicate persistent performance gaps between dominant and underrepresented languages, with little evidence that these disparities are narrowing. Consequently, current AI models remain trained and optimized for only a small fraction of the world's languages.
OBDILCI research also shows that the percentage of web content in English has decreased from 80 percent in 1998 to around 20 percent in 2009 which is a welcome development in the promotion of a multilingual cyberspace. There has to be deliberate and sustained efforts, especially by commercial technology companies to ensure the presence of diverse languages in cyberspace. To illustrate, Google Translate supports 249 languages and language varieties as of November 2024 (Google, 2024a). In June 2024, Google announced that 110 new languages had been added using AI (Google, 2024b). As of November 2024, Microsoft Translator supports 177 languages and language varieties (Microsoft, 2024).
Aligned with its mandate to preserve endangered languages and promote linguistic diversity, UNESCO supports the Language Technologies for All (LT4All). The program encompasses the development of simple tools facilitating cross-linguistic communication to complex machine translation systems which will enable access to information in native languages while fostering cross-cultural understanding LT4AII 2025). The need for interactive and engaging language learning applications and online resources as crucial elements in language preservation and development is well-recognized (Ajani, et.al, 2023).
The Austrian National Commission for UNESCO recently published the Documentation Framework for Linguistic Diversity which provides a standardized approach to describing and assessing the world's languages, with particular attention to minority, Indigenous, and endangered languages. The framework includes a matrix of descriptors covering linguistic facts, language status, language users, patterns of language use, and indicators of language dominance and endangerment. To demonstrate the application of these descriptors, the framework presents case studies from Mexico, Ghana, and Austria
(Austrian National Commission for UNESCO, 2026, p. 7).
AI and Media and Information Representation
Media and information representation refers to portrayal by media and information sources (inclusive of formats and platforms) of persons, places, events and objects by depicting, highlighting, or ignoring selected characteristics or properties. This framing can result in identities, images or perceptions of reality which can lead to stereotyping. Characteristics often represented are race, national or social origin, indigenous origin or identity, color, language, religion, gender identity, sexual orientation, political or other opinion, disability, and migrant or refugee status. Representation also applies to historical and current events and objects.
AI can reinforce and perpetuate biases or even inaccuracies when datasets available are used to reflect only perspectives from the main population (i.e., mainstream communities) because they had the advantage of populating the cyberworld with their perspectives early on while minorities have difficulties catching up in presenting their own perspectives. There can also be bias when datasets are based on historical data that embed biases. Another source of bias is when developers come from similar culture or sociopolitical background which can lead to creation of echo chambers (El-Guindy, 2022).
A cursory review of some historical and news narratives would reveal some words that may reflect a colonial perspective or mindset. For example, “discovery” of indigenous lands, conflict in “disputed” territories, and use of catchwords such as “illegal immigrants,” “pagan,” and “primitive.”
While AI is becoming an important tool in the newsrooms, the media’s responsibility remains to depict what is true about IP. The call for accuracy will be as strong as when the media focus on other subjects. Because AI can facilitate content production, the responsibility of ensuring the news content’s veracity and ensuring that the IP perspective is taken into account will greatly depend on the reporter’s skill, exposure to, and knowledge of the IP content.
In his presentation material entitled AI for Broadcast, Journalism, and Communication:
Staying Human in the Digital Age, multiplatform journalist Gajete (2024) tackled different ways that AI is affecting newsroom dynamics. From making sense of large data sets to generating visuals to support news materials, AI is now making it easier for media organizations to manage the processes that go into content production. However, Gajete stressed that “…the pace and extent of AI adoption will vary across organizations, influenced by factors such as technological capabilities and resource availability.”
In an interview article, Better Questions About Generative AI, published in Deloitte Insights Magazine Issue 33, 26 September 2024, Dr. Poornima Luthra, an associate professor at Copenhagen Business School proposes key questions to ask in pursuing a more inclusive cyberspace (countering AI bias) using her concept of active allyship: “Whose voice is missing? How is someone who’s different from myself, are they represented in what generative AI is providing us, in the responses that we’re getting? Is that perspective representative of the global majority? The Global South? The underrepresented, the underrecognized, the underestimated: Are those voices really there?”
AI offers opportunities to present a more balanced, fair, and accurate representation of indigenous peoples. However, there are some challenges in achieving this objective. First, AI content is generated from sources that are still written in colonial languages, which may perpetuate linguistic and cultural biases. Second, content (knowledge) that articulates or amplifies perspectives from indigenous communities, may still be lacking. Third, AI developers may lack cultural awareness and sensitivity. Fourth, indigenous peoples are not involved in designing, developing, and deploying AI systems. Fifth, AI systems may perpetuate biases and discrimination against indigenous peoples as developers’ biases can be embedded in AI systems.
Representation can impact an individual’s cognitive biases, a person’s construction of subjective reality. The person makes decisions not based on facts, rationality, and objective judgment.
AI literacy should be included in media and information literacy (MIL) to enable individuals to be sensitive to negative or undesirable representation, especially those who build pride and prejudice towards “different others.” Through MIL, individuals are able to recognize prejudice, hate, deception, or manipulation present in representation.
A Voice for the Voiceless
Diversity in media and information can refer to the presence of varied or differing “voices” and content (messages and images) representing geographic, cultural, religious, political, and socioeconomic variances. Diversity and pluralism are balancing efforts to counter so- called cultural homogenization which can erode distinct cultural identities.
Diversity and pluralism are pursued by providing distinct media platforms for ethnic communities and making existing and emerging platforms open, i.e., available and accessible, to the voices of all, especially the marginalized. UNDRIP (Articles 16.1 and 16.2) recognizes the right of IP to establish their own media in their own languages and to have access to all forms of non-indigenous media without discrimination. It also calls upon State- owned and privately owned media to duly reflect indigenous cultural diversity.
The lack or absence of diversity can lead to cultural homogenization with overdependence or unmindful use of algorithms, the building blocks of AI systems. Computer and data scientists and social scientists warn of the practice of technology companies to harvest and mine personal data/information that are used as raw materials in crafting customized messages to targeted audiences as bases for behavior change or adoption. Data is made available and accessible to technology companies every time their products and services are used, e.g., search engines, social media platforms, messaging apps, peer sharing platforms, etc. Part of this architecture is the use of algorithmic products designed to predict the behavior of technology users (Zuboff, 2019).
The danger of behavior modification leading to behavior homogenization with mainstream cultural audiences as standard is another threat minority audiences should be concerned with. Behavior modification can be a means or process toward behavior compliance, with powerful technology owners (mainstream communities) defining desired behavior standards and practices.
In addition to articulating or amplifying indigenous peoples’ voices, media, information and digital platforms are channels for intergenerational transmission of cultural values, cultural heritage, customs, and traditions. Intergenerational transmission facilitates
preservation of collective memory and cultural heritage. Can the existing AI ecosystem ensure that transmission occurs in a climate of free expression, tolerance, cultural integrity and authenticity, and transparency?
Among the major challenges in articulation, amplification, and intergenerational transmission is the participation of IP communities in all stages of the AI life cycle which depends on digital and AI literacy and of course, the openness of the system, especially to be challenged for inaccuracies and misrepresentations. AI life cycle covers research, design, and development to deployment and use, including disassembly and termination.
A related concept is authenticity which refers to faithfulness and accuracy to cultural traditions and customs. UNDRIP Article 31 stipulates that, “Indigenous peoples have the right to maintain, control, protect and develop their cultural heritage, traditional knowledge and traditional cultural expressions, as well as the manifestations of their sciences, technologies and cultures, including human and genetic resources, seeds, medicines, knowledge of the properties of fauna and flora, oral traditions, literatures, designs, sports and traditional games and visual and performing arts. They also have the 23 right to maintain, control, protect and develop their intellectual property over such cultural heritage, traditional knowledge, and traditional cultural expressions” (UNDRIP, 2007).
Intellectual Property, Data Sovereignty, and Open Access
Intellectual property encompasses creations of the human intellect that are protected by law against unauthorized use (Grizzle et al., 2021). In the context of artificial intelligence, governance frameworks should also respect Indigenous Peoples’ rights to their cultural heritage, traditional knowledge, and traditional cultural expressions, as recognized in Article 31(1) of the United Nations Declaration on the Rights of Indigenous Peoples.
When AI harvests and presents information, due diligence should be practiced to ensure respect for property rights of Indigenous communities and data sovereignty, i.e., Indigenous peoples’ data collection, storage, and use should be with consent. It is important to consider that the concept of (individual) property rights may be unknown to some indigenous communities. In some indigenous communities, communal (community) rights, especially those related to traditional knowledge (e.g., local arts, music, and literature), are recognized and need protection.
Intellectual property is balanced by open access. A media and information and AI literate individual should be able to claim free access to information and unrestricted use of electronic resources under the pivot of Open Access (and Open Science).
Access by IP to indigenous knowledge is a concern. In her article entitled Indigenous Knowledge and Archives: Accessing Hidden History and Understandings, Russell (2005) chronicled how access to knowledge resources is closely linked to identity and culture. Because indigenous families have “on occasion been denied access to their cultural knowledge,” they resorted to archives and libraries to search for historical information on their heritage. While some of the information may be available, certain conditions had to be met, such as when it concerns medical records. In one of the article’s case studies, Russell cited that records may be accessible if the patient was a family member or if the patient was born over 100 years ago. Still, having these reference materials on open access can pose a concern to families who might not want to have these records open for anybody to look at. What happens if these materials, some of which might be distressing for indigenous families, are made part of the AI ecosystem?
According to Jack McKenna, a writer for the open-access scientific publisher MDPI, AI could help address open access challenges such as the increasing volume of research data, language barriers, imbalances in research outputs among countries, and openness in datasets and content aggregators (McKenna, 2024). McKenna emphasized that “AI presents both a threat and opportunity. It requires reflection, adjustments, and adaptations… AI is changing Open Access; it’s changing everything” (McKenna, 2024).
In June 2024, UNESCO and The Royal Society convened the roundtable discussion Navigating the Intersection of Open Science and AI: Exploring Opportunities and Challenges in the Era of Artificial Intelligence (UNESCO & The Royal Society, 2024). Among the challenges discussed were the black-box nature of AI systems, the growing dominance of private companies over academic institutions in AI-powered research, the reproducibility crisis—whereby other researchers cannot replicate experiments conducted using AI tools— and data-related challenges, including issues of volume, heterogeneity, and potential bias (UNESCO & The Royal Society, 2024).
A status report on open access (OA) worldwide is available through the UNESCO Global Open
Access Portal (GOAP) (UNESCO, 2021). The portal provides country profiles, open access policies, repositories, mandates, key organizations, and other resources on the global state of open access.
Meanwhile, there are now thousands of online libraries and archives worldwide featuring digitized materials. The major online libraries are mostly maintained in western countries. These include World Digital Library (U.S. Library of Congress) Universal Digital Library, Project Guttenberg, and Google Books. IP communities have the right to know and access these archived materials to have a comprehensive account of their history and culture and to challenge possible inaccuracies and misrepresentations. Equally important are issues related to access and even ownership of physical archives stored in foreign lands. The inclusion of indigenous knowledge materials in online libraries and archives from IP communities themselves should be encouraged to have more balanced or accurate representation.
AI can reinforce and perpetuate biases or even inaccuracies when datasets available are used to reflect only perspectives from the main population (i.e., mainstream communities) because they had the advantage of populating the cyberworld with their perspectives early on while minorities have difficulties catching up in presenting their own perspectives. There can also be bias when datasets are based on historical or archival data that embed biases.
Decolonizing Archives
AI may either challenge or reinforce colonialism when the dominant narratives, training data, and generated content reflect colonial biases.
Scholars and researchers should examine how to “decolonize” global archiving system to promote access to information. This involves actual access (offline and online) of archival materials in Western libraries and archives, digital repatriation, restitution of cultural heritage materials, reparations and compensation for the removal and exploitation of historical and cultural heritage materials.
Some archival documents can be sources of 'original' disinformation, or 'fake news,' as they were written from a colonial perspective, reflecting biased or distorted views of historical events. Access to archival documents housed in Western libraries and archives is essential for
scholars from formerly colonized countries to verify, validate, and potentially correct these records.
Filipino film scholar-historian Nick Deocampo continues to pursue his passion for researching film history, including the search for original copies of pioneering Asian films. His book, Asia’s Lost Films (2006), documents the discovery of at least 30 lost films. He has visited archives, libraries, and private collections around the world in search of these lost cultural treasures. Through this work, he aims to rewrite the history of Asian cinema by challenging dominant narratives that have marginalized the region’s rich cinematic diversity.
Cultural Justice
AI development should at all times recognize the rights and interests of indigenous communities particularly in a multicultural/multiethnic setting where the culture of the majority usually predominates. Cultural justice means values, beliefs, traditions, bodies of knowledge of ethnic minorities are adequately protected. Cultural justice also means indigenous communities, as rights holders, can demand from the State and other duty bearers, the respect and fulfillment of their rights under a regime of equality and freedom.
Pope Leo XIV’s Magnifica Humanitas extensively discusses the threat that a digital ecosystem controlled by technological and economic elites poses to social justice, cultural justice, and the pursuit of the common good (Pope Leo XIV, 2026). To quote the encyclical: “In many cases within the digital context, control over platforms, infrastructure, data and computing power does not rest with States, but with major economic and technological actors. These entities effectively set the conditions for access, determine the rules of visibility and shape the very possibilities for participation. When such power is concentrated in the hands of a few, it tends to become opaque and evade public oversight, increasing the risk of distorted forms of development that give rise to new dependencies, exclusions, manipulations and inequalities” (Pope Leo XIV, 2026, para. 95).
One of the most important recent documents protecting the rights of indigenous peoples is the Report of the Working Group on the Issue of Human Rights and Transnational Corporations and Other Business Enterprises entitled, Guidance on The Right of Indigenous Peoples to Free, Prior
and Informed Consent (FPIC) in The Context of Business Activities. The Report was issued during the Human Rights Council 62nd session from 15 June–10 July 2026.
The Report’s annex defines FPIC as the right of indigenous peoples to decide, through their own institutions and decision-making systems, whether and under what conditions a project affecting their rights, including to lands, territories, resources, and cultural integrity, or future may proceed. Although the Report acknowledges that FPIC has advanced in national laws and international human rights standards, implementation remains inconsistent.
The Guidance (2026) clarifies the responsibilities of governments and businesses in protecting and respecting Indigenous Peoples' right to FPIC when business activities may affect their rights, ancestral lands, territories, natural resources, livelihoods, or cultural heritage.
Socio-Technical Approach to AI Development
In Harnessing AI in Developing Countries: Why a Socio-Technical Approach Matters, Lallana (2024) pointed out that AI systems are not just algorithms, but that these have social elements, too, that are essential to the full human understanding of how these systems work. Citing NIST (2022), he emphasized that a socio-technical approach “takes into account the values and behavior modeled from the datasets, the humans who interact with them, and the complex organizational factors that go into their commission, design, development, and ultimate deployment.”
This approach enables individuals to understand that AI is part of a larger web of processes and social systems and that it is marked by the values and interests of its developers. Thus, this may tend to further marginalize the already marginalized communities especially since they are not involved in the development of these AI systems.
Lallana posits the importance of AI governance in the context of the sociotechnical approach. He cited a suggestion of a Wilson Center study on Governing AI to focus on algorithms, data, actors, and outcomes to fully determine the intent and purpose, as well as the behavioural norms that can be supported.
He highlighted public interest as another area for consideration in AI governance, adding that public contestability should be a vital component to enable individuals to “collectively contest decisions to ensure that public interest is served.” This important feature can also be invoked
by IP communities who feel the need to correct information about their cultural and historical heritage found in open access, thus helping ensure the accuracy of the information that can be accessed freely.
This socio-technical approach finds support in Magnifica Humanitas, which proposes that the power of digital infrastructures and algorithms should be assessed according to whether they truly foster participation and responsibility, protect the vulnerable, ensure fair access to opportunities, and remain directed toward the common good” (Pope Leo XIV, 2026, para. 9, pages 36-37).
The Myanmar Experience
In the recent past, UNESCO Yangon Office developed toolkits and standards intended for multiplatform content producers from mainstream and indigenous communities, and therefore, relevant to AI producers. These toolkits and standards can serve as models in addressing some of the issues mentioned earlier, and in ensuring that cultural protocols and ethical considerations are upheld in content development, sharing, and exchange. These toolkits commit to the enjoyment and exercise of all media freedoms including freedom of expression, freedom of the press, and freedom of information.
These toolkits include Myanmar Journalists’ Toolkit in Culture and Ethnic Reporting; UNESCO Myanmar Recommendations for Religion and Culture- sensitive Performances/Presentations/Ceremonies; UNESCO Myanmar Recommendations on the Use of Religious and Cultural (Community) Artifacts and Symbols; and UNESCO Myanmar Recommendations for Religion and Culture-sensitive Multimedia Productions.
The Myanmar Journalist Toolkit covers the following topics: Ethno-Cultural Landscape of Myanmar; Ethnic Media in Myanmar; Why Cultural and Ethnic Reporting is Important; Do’s and Don’ts in Cultural and Ethnic Reporting; Gender Sensitivity in Cultural and Ethnic Reporting; Safety and Protection of Journalists Covering Ethnic Communities; and Story Ideas.
Meanwhile the UNESCO Recommendations completed in 2022, underwent validation and verification by local culture experts, ethnic and monastic educators, and officers and staff from
a relevant government agency (e.g., Ministry of Ethnic Affairs).
The following are some examples of proposed guidelines:
• Understanding, appreciation, and respect for objectives, meanings, and history of performance, presentation, and ceremony;
• Respect for authenticity of elements of performance, presentation, and ceremony (e.g., music, instruments, attire, and movements);
• Use of the word religious, indigenous/ethnic traditional dress/attire instead of costume when referring to indigenous clothing or garments;
• Avoidance of commercializing cultural and religious artifacts and symbols (including modifications that may change the essence of the artifacts and symbols); and
• Consistency in the documentation of traditional and religious adornments, ornaments, speech, and locations (environment) consistent with historical timelines.
The Pax Silica Project as Test Case in AI Planning
The proposed AI Hub in New Clark City, Philippines, under the United States’ Pax Silica Initiative represents an important test case for rights-centered AI planning and governance. Pax Silica is the US Department of State’s flagship effort on AI and supply chain security. Twenty three countries , including the Philippines, has so far signed the Pax Silica Declaration ((U.S. Department of State, n.d.). The Philippines joined Pax Silica on 16 April 2026.
In the Philippines, the proposed AI Hub covers approximately 4,000 acres, the proposed development is expected to include AI computing infrastructure, advanced manufacturing facilities, AI hardware production, AI training centers, and other related facilities. Its proponents anticipate several benefits, including the creation of high-value jobs, technology transfer, increased foreign investment, and the development of local AI capabilities.
However, Indigenous communities, particularly the Aeta people, have raised sociocultural concerns about the proposed project, expressing caution and, in some cases, outright opposition. A major concern is the potential threat to their ancestral domains, including the possible loss of cultural heritage sites and ancestral lands that they have traditionally occupied
and cultivated. For the Aeta, land is not merely an economic asset; it is integral to their identity, history, culture, and collective existence.
The proposed development may also threaten traditional livelihoods and Indigenous knowledge systems, particularly those sustained through agriculture and transmitted across generations. Aeta leaders have expressed concern about the potential disruption of intergenerational cultural continuity. Critics argue that projects affecting Indigenous ancestral domains should not proceed without securing Free, Prior and Informed Consent (FPIC), consistent with applicable legal frameworks and international standards on Indigenous Peoples’ rights.
Rights-Based Approach to AI Development and Regulation
Recalling the communication rights of Indigenous Peoples is aligned with the rights-based approach to AI regulation discussed in the Consultation Paper on AI Regulation (UNESCO, 2024). The paper identifies the rights-based approach as one of the nine emerging approaches to AI regulation (UNESCO, 2024).
This approach posits that “regulations are justified to protect rights, promote distributional justice, and further social objectives… establishes mandatory rules to guarantee the respect, protection, and promotion of rights, including human rights and other economic or social rights, along the AI system’s life cycle.”
The paper also recalls the framework proposed by Margaret Mitchell, which identifies the different stakeholder groups that could be negatively impacted at each stage of the AI system life cycle—including data creators, data subjects, AI developers, AI deployers, AI users, and AI- affected people—the human rights that could be affected in each case, and the corresponding regulatory measures needed to protect those rights (UNESCO, 2024, p. 37).
The paper cited as examples some provisions of the Brazilian Bill No. 2238 / 2023 which “enshrines rights for persons affected by AI systems (Articles 5 and 6), rights associated with ‘information and understanding of the decisions made by artificial intelligence systems’ (Articles 7 and 8), rights to ‘challenge decisions and request human intervention’ (Articles 9 and 11), and the ‘right to non-discrimination and the correction of direct discriminatory biases, indirect, illegal or abusive’ (Article 12)” (underscoring provided).
Policy and Action Agenda
The following are proposed policy and action agenda toward an IP human right-centered AI:
1. Recognize and integrate rights of Indigenous peoples in global, national, and local AI (regulatory) policies and development plans/programs.
2. Adapt the Human Rights Council Working Group’s Report on Guidance on The Right of Indigenous Peoples to Free, Prior and Informed Consent (FPIC) in The Context of Business Activities.
3. Adapt into local context the UNESCO Recommendation on the Ethics of Artificial Intelligence.
4. Expand access to reliable and affordable technology and connectivity in IP communities.
5. Sustain investment in Language Technologies for All (LT4All) by State, business and industry, NGOs, etc.
6. Provide AI literacy, as component of MIL, to indigenous peoples.
7. Extend MIL education, especially culture sensitivity, to AI developers.
8. Advance MIL education to enable IP to demand transparency, responsibility, and accountability from technology companies including exacting data challenges and corrections.
9. Ensure the participation of Indigenous peoples in the AI life cycle, i.e., designing, developing, and deploying AI systems.
10. Support and sustain the presence of Indigenous knowledge, including history, culture, and heritage in the cyberspace.
11. Use AI to document, translate, and revitalize Indigenous languages.
12. Promote multilingualism in cyberspace including engagement by technology companies.
13. Respect for intellectual and communal property rights of Indigenous peoples.
14. Prioritize Indigenous data sovereignty and governance including Indigenous data governance, ownership, and control.
15. Support open (science) access by indigenous peoples as mechanism for free access to information and unrestricted use of electronic resources .
16. Adequate representation of IP archives in major global online libraries.
17. Establish alliance with global indigenous AI research and development and advocacy groups.
References
Ajani, Y., Oladokun, B., Olarongbe, S., Amaechi, M., Rabiu, N., & Bashorun, M. (2024). Revitalizing indigenous knowledge systems via digital media technologies for sustainability of indigenous languages. Preservation, Digital Technology & Culture, 53(1), 35-44. doi: 10.1515/pdtc-2023-0051
Austrian Commission for UNESCO. (Ed.). (2026). Documentation framework for linguistic diversity. Austrian Commission for UNESCO. https://doi.org/10.5281/zenodo.18618404
Berlin Declaration on Open Access to Knowledge in the Sciences and Humanities. (n.d.). https://openaccess.mpg.de/Berlin-Declaration
Deocampo, N. (Ed.). (2006). Lost films of Asia. Anvil Publishing.
El-Guindy, M. (2022). Malicious use of artificial intelligence. Legal and ethical implications. [PowerPoint presentation]. AMIC Webinar on AI.
Gajete, S. (2024). AI for broadcast, journalism, and communication: Staying human in the digital age. [PowerPoint presentation]. Presentation during the webinar Breaking Waves: Navigating AI IN Philippine Broadcasting convened by the Department of Broadcasting and Telecommunications, College of Development Communication.
Grizzle, A., Wilson, C., Tuazon, R. R., Cheung, C.K., Lau, J., Fisher, R., Gordon, D., Akyempong, K., Singh, J., Carr, P.R., Stewart, K., Tayie, S., Suraj, O., Jaakola, M., Thésée, G., Gulston, C., Andzsongo Menyeng, B.P., & Zibi Fama, P.A. (2021). Media and information literate citizens: Think critically, click wisely. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000377068
https://unesdoc.unesco.org/ark:/48223/pf0000388350
https://www.ohchr.org/en/professionalinterest/pages/ccpr.aspx
International Covenant on Civil and Political Rights. (n.d.).
International Telecommunications Union. (2023, November). Digital inclusion of all. https://www.itu.int/en/mediacentre/backgrounders/Pages/digital-inclusion-of-all.aspx
Ireton, C., & Posetti, J. (Eds.). (2018). Journalism, 'fake news' & disinformation: Handbook for journalism education and training. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000265552_eng
Kurtz, A., & Blau, A. (2024, September 26). Better questions about generative AI. Deloitte Insights Magazine, 33. https://www2.deloitte.com/us/en/insights/topics/emerging- technologies/better-questions-about-generative-ai.html
Lallana, E. (2024). AI in developing countries: Why a sociotechnical approach matters.
Languages. https://www.unesco.org/en/articles/unesco-hosts-ad-hoc-expert- meeting-world-atlas-languages?hub=67103
Magnifica Humanitas. (2025). Magnifica Humanitas. Libreria Editrice Vaticana. https://www.vatican.va/content/leo-xiv/en/encyclicals/documents/20260515-magnifica- humanitas.html
Maraña, M. (2010). Culture and development; evolution and prospects. UNESCO Etxea Working Papers, 1. https://www.ehu.eus/documents/3120344/3356415/Culture_Development.pdf/837c2fac- 4627-4a29-a2fica 1b-f41ddbf52ab2
McKenna, J. (2024, October 15). Why open access is important. MDPI. https://blog.mdpi.com/2024/10/15/open-access-is- important/?utm_source=facebook&utm_medium=social_corp&utm_campaign=majorcampaign_ oa w24_blog_importance&fbclid=IwY2xjawGShj5leHRuA2FlbQIxMAABHaLKGRF3kOKjQbv6TS89reAS dKx 1jydAPQexg0Cs8h2Om2nMjzAq0Tpx0g_aem MZorUTscRGyLSUqkNbhrg
Pimienta, D., Prado, D., & Blanco, Á. (2009). Twelve years of measuring linguistic diversity in the internet: Balance and perspectives. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000187016
Reparejo, A. J. R. (2026, April 23). BCDA eyes AI industrial hub in New Clark City under US-led Pax Silica initiative. Philippine Information Agency. https://pia.gov.ph/news/bcda-eyes-ai-industrial-hub-in- new-clark-city-under-us-led-pax-silica-initiative/
Reuters. (2026, April 17). Philippines, US to build industrial hub to strengthen supply chain security. https://www.reuters.com/world/asia-pacific/philippines-us-build-industrial-hub-supply-chain- security-2026-04-17/
Russell, L. (2005). Indigenous knowledge and archives: Accessing hidden history and understandings. Australian Academic & Research Libraries, 36:2, 161-171, DOI: 10.1080/00048623.2005.10721256
The 1954 Hague Convention for the Protection of Cultural Property in the Event of Armed Conflict. (1954). https://en.unesco.org/protecting-heritage/convention-and-protocols/1954- convention#:~:text=The%201954%20Hague%20Convention%20aims,regardless%20of%20their%2 0o rigin%20or
Tuazon, R. R. (2023). Myanmar journalists’ toolkit in culture and ethnic reporting.
Tuazon, R. R. & Myat, Mon Mon. (2023). Myanmar journalists’ toolkit in culture and ethnic reporting. https://unesdoc.unesco.org/ark:/48223/pf0000388350
U.S. Department of State. (n.d.). Pax Silica. https://www.state.gov/pax-silica?utm_source
UNESCO Myanmar Office. (2023). UNESCO Myanmar Recommendations for Religion and Culture- sensitive Performances/Presentations/Ceremonies; UNESCO Myanmar Recommendations on the Use of Religious and Cultural (Community) Artifacts and Symbols; and UNESCO Myanmar Recommendations for Religion and Culture-sensitive Multimedia Productions
UNESCO Universal Declaration on Cultural Diversity. (2001). http://portal.unesco.org/en/ev.php- URL_ID=13179&URL_DO=DO_TOPIC&URL_SECTION=201.html
UNESCO. (2005). Convention on the protection and promotion of the diversity of cultural expressions. passeport-convention2005-web2.pdf (unesco.org)
UNESCO. (2021). Recommendation on the ethics of artificial intelligence. https://unesdoc.unesco.org/ark:/48223/pf0000381137
UNESCO. (2022). ReShaping policies for creativity: Addressing culture as a public good. https://www.unesco.org/reports/reshaping- creativity/2022/en/download-report
UNESCO. (2024). Consultation paper on AI regulation: Emerging approaches across the world. https://unesdoc.unesco.org/ark:/48223/pf0000390979
UNESCO. (2024). Exploring opportunities and challenges at the intersection of open science and artificial intelligence. https://www.unesco.org/en/articles/exploring-opportunities-and- challenges- intersection-open-science-and-artificial-intelligence
UNESCO. (2024). Language technologies for all – LT4All 2025. Advancing humanism through language technologies. https://www.unesco.org/en/articles/language-technologies-all-lt4all- 2025
UNESCO. (2024). UNESCO hosts the Ad-Hoc expert meeting on World Atlas of
United Nations Declarations on Human Rights. (1948). https://www.ohchr.org/EN/UDHR/Documents/UDHR_Translations/eng.pdf
United Nations Declarations on the Rights of Indigenous Peoples. (2007). https://www.un.org/development/desa/indigenouspeoples/wp- content/uploads/sites/19/2018/11/UNDRIP_E_web.pdf
United Nations. (2024). Summit of the future outcome documents. https://www.un.org/sites/un2.un.org/files/sotf-pact_for_the_future_adopted.pdf
United Nations. (2026). International Day of the World's Indigenous Peoples: Background. https://www.un.org/en/observances/indigenous-day
United Nations. (2026). Preliminary Report of The Independent International Scientific Panel on AI: Evidence-Based Assessment Of Opportunities, Risks and Impacts of Artificial Intelligence. United Nations. https://www.un.org/independent-international-scientific-panel- ai/sites/default/files/2026-07/en_Preliminary%20Report_.pdf
Villanueva, Joann. (23 July 2026). Planned Pax Silica project heeds national, international laws: BCDA. Philippine News Agency. https://www.pna.gov.ph/articles/1280187?utm_source=chatgpt.com
Working Group on the Issue of Human Rights and Transnational Corporations and Other Business Enterprises. (2026). Guidance on the right of Indigenous Peoples to free, prior and informed consent in the context of business activities (A/HRC/62/36/Add.2). United Nations Human Rights Council
World Bank. (2025). Indigenous Peoples. https://www.worldbank.org/en/topic/indigenouspeoples
Zuboff, S. (2019). The age of surveillance capitalism: The fight for a human future at the new frontier of power. BBS PublicAffairs.
