Digital Empowerment Foundation

The Digital Empowerment Foundation (DEF), drawing from over two decades of grassroots experience bridging digital divides across India, believes the first Global Dialogue on AI Governance must deliver outcomes that center the unconnected, the underserved, and the underrepresented. Success should be measured not by declarations alone, but by tangible commitments that ensure AI serves humanity equitably. 1. Inclusive and Equitable Participation A successful Dialogue must institutionalize meaningful participation of civil society, grassroots communities, indigenous peoples, women, persons with disabilities, and actors from the Global South, not as observers, but as co-architects of governance frameworks. Multilingual access, accessible formats, and dedicated funding for community representation are non-negotiable. 2. A Human-Centered Phygital Framework The Dialogue should reaffirm that AI governance must be rooted in international human rights law, dignity, and social justice. Outcomes must explicitly address algorithmic discrimination, surveillance harms, labour displacement, linguistic exclusion, and impacts on marginalized communities, including informal workers and rural populations. 3. Bridging the AI Divide Success requires concrete pathways to prevent a widening "AI divide." This includes commitments to open datasets in underrepresented languages, public digital infrastructure, equitable compute access, capacity-building for the Global South, and support for community-led and frugal AI innovations that solve local development challenges. 4. Accountability and Redress Mechanisms The Dialogue should establish clear principles for transparency, explainability, independent auditing, and accessible grievance redressal, particularly for communities affected by automated decision-making in welfare, credit, policing, and public services. 5. A Permanent, Inclusive Follow-up Architecture Finally, success means agreeing on a sustained, transparent, and participatory institutional mechanism, linked to the Global Digital Compact and Scientific Panel on AI—that ensures continuous multistakeholder engagement, with civil society and grassroots voices structurally embedded. For DEF, success means AI governance 'with' communities, not merely 'for' them.

• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights
• Open-source software, open data and open AI models

The Digital Empowerment Foundation (DEF) has spent over two decades working at the last mile of the digital divide, embedded in more than 2400 of India's most underserved locations. Our selection reflects ground-realities witnessed daily in rural, tribal, and marginalized communities across the country. AI Capacity-Building is our foremost priority because the AI revolution risks bypassing entire populations. Without grassroots digital and AI literacy, training of community intermediaries, equitable compute access, and institutional capacity in the Global South, governance frameworks will remain inaccessible to those who need protection most. Capacity is the prerequisite for meaningful participation in AI governance itself. Social, Economic, Ethical, Cultural, and Linguistic Implications matter urgently because AI is already reshaping welfare delivery, livelihoods, and cultural representation. In India alone, hundreds of languages remain invisible to AI systems, informal workers face displacement without safety nets, and algorithmic decisions exclude eligible beneficiaries from entitlements. These multidimensional harms demand integrated governance responses, not siloed technical fixes. Protection and Promotion of Human Rights anchors all credible AI governance. DEF documents how automated systems in identity, policing, credit scoring, and welfare disproportionately harm the poor, women, Dalits, Adivasis, religious minorities, and persons with disabilities. Embedding international human rights law-dignity, privacy, non-discrimination, freedom of expression, and access to remedy, is non-negotiable for legitimate AI governance. Open-Source Software, Open Data, and Open AI Models are essential to counter monopolistic concentration of AI power. Openness enables community innovators, public-interest researchers, and Global South developers to build locally relevant, linguistically inclusive, and frugal AI solutions. Open ecosystems, paired with public digital infrastructure, democratize AI and prevent a new form of digital colonialism. These priorities ensure AI governance is inclusive, rights-based, community-rooted, and equity-driven, closing rather than widening the global AI divide.

While the seven thematic areas in Resolution 79/325 are comprehensive, the Digital Empowerment Foundation (DEF) believes several cross-cutting and emerging issues require explicit attention to ensure AI governance is truly inclusive and future-ready: 1. Environmental and Ecological Sustainability of AI The exponential energy, water, and mineral footprint of AI infrastructure-data centres, training compute, and e-waste-poses severe climate and ecological risks, disproportionately affecting Global South communities living near extraction and disposal sites. AI governance must integrate planetary boundaries and just transition principles. 2. AI and the Future of Work, Especially Informal Labour Most global labour is informal, particularly in the Global South. Discussions on AI and work overwhelmingly focus on formal employment, ignoring gig workers, home-based workers, artisans, and agricultural labourers whose livelihoods are being reshaped by algorithmic management and automation without social protection. 3. Gendered and Intersectional Harms of AI Algorithmic bias intersects with gender, caste, race, disability, sexuality, and migration status in compounding ways. A dedicated intersectional lens-beyond generic "non-discrimination"-is essential, including addressing AI-enabled gender-based violence, deepfakes, and online harassment. 4. AI in Conflict, Surveillance, and Democratic Integrity The use of AI in mass surveillance, predictive policing, autonomous weapons, border control, and election manipulation threatens democratic freedoms and civic space. These risks deserve urgent, dedicated governance attention. 5. Children, Youth, and Future Generations AI's impact on cognitive development, education, mental health, and child safety online warrants specific safeguards and intergenerational equity principles. 6. Indigenous Data Sovereignty and Cultural Rights Indigenous and traditional knowledge systems risk extraction and misuse by AI training pipelines. Governance must recognize collective data rights, community consent, and cultural sovereignty. 7. Meaningful Connectivity as a Precondition Nearly 2.6 billion people remain offline. Without universal, meaningful, and affordable connectivity, AI governance debates exclude those most affected.

The Digital Empowerment Foundation (DEF) observes that governance gaps in our four priority areas are producing tangible harms and missed opportunities across India and the broader Global South. Challenges: 1. Capacity-Building Deficit: Despite India's digital ambitions, frontline communities, rural women, Adivasis, Dalits, persons with disabilities, and informal workers remain excluded from AI literacy and skilling. Public institutions, panchayats, and civil society lack the technical capacity to audit or contest AI systems deployed in welfare and governance. This creates a widening "AI divide" within and between nations. 2. Socio-Economic and Linguistic Exclusion: Algorithmic systems in welfare delivery (e.g., ration, pensions, MGNREGA wages) have wrongfully excluded eligible beneficiaries, with documented cases of starvation deaths linked to biometric and automated failures. India's 700+ languages and oral traditions are largely absent from mainstream AI models, erasing cultural and linguistic diversity. Informal workers face algorithmic wage-setting and surveillance without legal protection. 3. Human Rights Erosion: Expanding deployment of facial recognition, predictive policing, and AI-enabled surveillance, often without legal safeguards threatens privacy, dissent, and minority rights. Deepfakes and AI-generated disinformation increasingly target women, journalists, and marginalized groups, with weak redressal mechanisms. 4. Closed AI Ecosystems: Concentration of AI capabilities among a few global corporations limits sovereign and community innovation. High costs of compute, proprietary datasets, and restrictive licensing prevent public-interest and Global South developers from building locally relevant solutions. Opportunities: DEF's two decades of last-mile work demonstrate that community-led, frugal, and rights-based AI is achievable. Our network of 2400+ Community Information Resource Centres (CIRCs), Soochnapreneurs (rural information entrepreneurs), and Digital Foot Soldiers offers a ready grassroots infrastructure to democratize AI literacy, demystify algorithms, and build community resilience against algorithmic harms. India's linguistic diversity, vibrant civil society, and open-source ecosystem create fertile ground for community-owned datasets, indigenous-language AI models, and public-interest applications addressing livelihoods, agriculture, health, and education. Strengthening multistakeholder governance, anchored in openness, human rights, and grassroots capacity, can transform marginalized communities from passive recipients into co-creators of AI, ensuring technology advances dignity, equity, and sustainable development across the Global South.

The Digital Empowerment Foundation (DEF) believes the Global Dialogue on AI Governance can serve as a transformative multilateral platform, not merely a forum for declarations, but a catalyst for inclusive, equitable, and action-oriented international cooperation. Its role should be defined by the following functions: 1. Bridging Governance Fragmentation Today, AI governance is fragmented across regional frameworks (EU AI Act, ASEAN guidelines, African Union strategy), national laws, and industry self-regulation. The Dialogue can foster interoperability of governance approaches, enabling coherence without imposing uniformity, respecting sovereignty while preventing regulatory arbitrage and a race to the bottom. 2. Amplifying Global South and Grassroots Voices The Dialogue must rebalance a conversation currently dominated by a few governments and corporations. It can institutionalize structural participation of civil society, grassroots actors, indigenous communities, women, and Global South stakeholders, ensuring AI governance reflects lived realities, not just boardroom priorities. 3. Anchoring AI in Universal Values As a UN-led process, the Dialogue uniquely carries the legitimacy to anchor AI governance in international human rights law, the Sustainable Development Goals, and the Global Digital Compact. This universal normative grounding can prevent AI governance from being reduced to narrow geopolitical or commercial interests. 4. Enabling Equitable Capacity and Knowledge Sharing The Dialogue can catalyze South-South and triangular cooperation, sharing open datasets, models, compute resources, technical expertise, and grassroots innovations. It can support pooled funding mechanisms and public goods that close the global AI divide. 5. Building Trust, Accountability, and Redress Through transparent reporting, independent scientific input (linked to the Scientific Panel on AI), and inclusive follow-up architecture, the Dialogue can build mutual trust among states, communities, and industry, essential for collective action on shared risks like disinformation, surveillance, and existential AI harms. The Dialogue's success lies in transforming international cooperation from rhetoric into rights-based, inclusive, and tangible action.

The Digital Empowerment Foundation (DEF) recommends that the Global Dialogue on AI Governance build upon a rich ecosystem of existing initiatives, particularly those that center grassroots communities, digital rights, and the Global South, while filling critical gaps in inclusivity, equity, and last-mile accountability. Key Initiatives and Mechanisms to Build Upon 1. UN-Led Frameworks - Global Digital Compact (GDC) and the forthcoming Scientific Panel on AI as foundational anchors - UNESCO Recommendation on the Ethics of AI and its Readiness Assessment Methodology - Internet Governance Forum (IGF) and its national/regional chapters (including IGF India, where DEF actively engages) - ITU's AI for Good and WSIS+20 review processes - OHCHR B-Tech Project on human rights and digital technologies 2. Intergovernmental and Regional Frameworks - OECD AI Principles and the Global Partnership on AI (GPAI), which India currently chairs - Council of Europe Framework Convention on AI - African Union Continental AI Strategy, ASEAN Guide on AI Governance, and EU AI Act as reference points - G20 Digital Economy Working Group and G77 + China South-South cooperation tracks 3. Civil Society and Digital Rights Coalitions - Association for Progressive Communications (APC), Global Digital Inclusion Partnership (GDIP), Access Now / RightsCon, and the Freedom Online Coalition - Dynamic Coalitions at IGF on community connectivity, gender, and digital rights—where DEF has long contributed - Grassroots networks like the Community Networks movement and Digital Empowerment Foundation's own CIRCs, Soochnapreneurs, and digital foot soldiers enabling last-mile inclusion. The Dialogue can uniquely deliver: (i) UN-anchored legitimacy transcending narrow geopolitical interests; (ii) structural inclusion of grassroots, women, indigenous, and Global South voices; (iii) harmonization without homogenization of fragmented frameworks; (iv) rights-based, SDG-aligned normative grounding; and (v) equitable capacity, open resources, and redress mechanisms, translating global principles into last-mile community impact.

The Digital Empowerment Foundation (DEF) believes the Global Dialogue on AI Governance must be designed as a truly multistakeholder, inclusive, and action-oriented process—drawing on the proven model of the Internet Governance Forum (IGF) while addressing its limitations through stronger grassroots representation and concrete deliverables. Stakeholder Contributions 1. Member States should share national AI strategies, regulatory experiences, and commit to equitable resource-sharing, capacity-building, and human rights compliance. 2. UN Agencies (UNESCO, ITU, UNDP, OHCHR) should provide technical expertise, normative frameworks, and coordination across the UN system. 3. Civil Society and Grassroots Organizations like Digital Empowerment Foundation should bring lived community realities, document algorithmic harms, advocate for digital rights, and represent voices of women, indigenous peoples, persons with disabilities, and informal workers. 4. Academia and Research Institutions should contribute independent evidence, impact assessments, and interdisciplinary scholarship—particularly from Global South institutions. 5. Private Sector (including SMEs and startups, not only Big Tech) should commit to transparency, human rights due diligence, open standards, and responsible innovation. 6. Technical Community and Open-Source Coalitions should advance interoperability, open models, and public-interest digital infrastructure. 7. Youth, Women, and Indigenous Networks must be structurally embedded, not tokenized. Recommended Format and Structure 1. Hybrid and Multilingual Participation: Ensure in-person, virtual, and regional satellite events with interpretation in UN and major regional languages, accessible formats for persons with disabilities, and dedicated funding for Global South and grassroots travel. 2. Tiered Architecture: Combine high-level plenaries, thematic working groups, regional consultations, and community-level dialogues that feed grassroots inputs upward. 3. Open and Transparent Inputs: Public calls for written submissions, open documentation, and live-streamed sessions. 4. Action-Oriented Outputs: Concrete roadmaps, measurable commitments, and follow-up mechanisms—not only declarations. 5. Permanent Inclusive Secretariat: With civil society and Global South co-leadership, ensuring continuity, accountability, and equitable agenda-setting.

The Digital Empowerment Foundation (DEF) observes that current global AI governance discussions remain dominated by a narrow set of governments, corporations, and elite technical experts—largely from the Global North. This exclusion undermines the legitimacy, relevance, and equity of emerging frameworks. Underrepresented Voices and Communities 1. Grassroots and Last-Mile Communities: Rural populations, informal workers, smallholder farmers, artisans, and gig workers whose lives are reshaped by AI but who are absent from policy tables. 2. Global South Civil Society and Governments: Particularly from Africa, South Asia, Latin America, Small Island Developing States (SIDS), and Least Developed Countries (LDCs). 3. Women and Gender-Diverse Communities: Especially those facing intersectional discrimination, AI-enabled gender-based violence, and deepfake harms. 4. Indigenous Peoples and Traditional Knowledge Holders: Whose data, languages, and cultural heritage are extracted by AI systems without consent. 5. Linguistic Minorities: Speakers of low-resource languages—India's 700+ languages, African languages, and oral traditions—are largely invisible to mainstream AI. 6. Persons with Disabilities: Often excluded from both AI design and governance discussions despite being deeply affected by accessibility gaps. 7. Children, Youth, and Future Generations: Whose cognitive, educational, and rights-based interests are inadequately represented. 8. Dalits, Adivasis, Religious Minorities, Refugees, and Stateless Persons: Communities facing compounded algorithmic discrimination. 9. Frontline Public Service Workers: Teachers, health workers, panchayat members, and welfare officers navigating AI deployment daily. Pathways to Inclusion 1. Dedicated Funding and Fellowships for grassroots, women, indigenous, and Global South participation. 2. Structural Quotas and Co-Leadership Roles in working groups and secretariats—not tokenistic observer status. 3. Multilingual, Accessible, Hybrid Formats with interpretation, sign language, and disability-inclusive design. 4. Community-Level Consultations feeding grassroots inputs into global processes through trusted intermediaries like DEF's CIRCs and Soochnapreneurs. 6. Indigenous Data Sovereignty Protocols recognizing collective consent and cultural rights. Meaningful inclusion transforms communities from subjects of AI governance into co-architects.

The Digital Empowerment Foundation (DEF) believes that the format of the Global Dialogue on AI Governance is as critical as its content. Traditional plenaries and panel-driven formats often exclude grassroots voices and reduce dialogue to performative exchanges. Drawing on DEF's two decades of grassroots engagement, we recommend the following innovative, inclusive, and dynamic formats: 1. Community-Anchored "Reverse Dialogues" Instead of experts speaking about affected communities, host sessions where grassroots actors, informal workers, indigenous leaders, and women from rural areas lead the conversation, with policymakers and technologists as listeners. DEF's Soochnapreneur and Digital Foot Soldier networks can anchor such sessions from the last mile. 2. Distributed Regional and Local Hubs Create satellite dialogues in villages, community networks, universities, and civil society spaces across the Global South, feeding inputs upward through a structured rapporteur system—ensuring geographic and socio-economic diversity. 3. Living Case Studies and Algorithmic Harm Documentation Curate real-life testimonies and evidence-based case studies of algorithmic exclusion (e.g., welfare denials, biometric failures, gendered deepfakes), making harms visible and human rather than abstract. 4. Multilingual and Multimodal Storytelling Use video, podcasts, community radio, illustrated narratives, and oral traditions to capture perspectives from non-literate, non-English-speaking, and indigenous communities, transcending text-heavy diplomatic formats. 5. Policy Hackathons and Co-Design Labs Bring together civil society, technologists, government, and affected communities to co-create solutions, open datasets, accountability tools, redress mechanisms, rather than only debating principles. 6. Youth and Intergenerational Dialogues Dedicated tracks where youth, students, and first-time participants engage with veteran policymakers, ensuring intergenerational equity and fresh perspectives. 7. Open Digital Commons Platform A persistent, multilingual, accessible online platform for continuous input, deliberation, voting on priorities, and transparent documentation, extending the Dialogue beyond episodic meetings. 8. Accountability Town Halls Annual public town halls where stakeholders report progress against commitments, ensuring dialogue translates into measurable action.

The Digital Empowerment Foundation (DEF), through over two decades of last-mile digital empowerment work, offers the following community-rooted practices and platforms as concrete approaches to inclusive AI governance, demonstrating that rights-based, equitable AI is achievable when grounded in grassroots realities. DEF's Grassroots Practices and Platforms 1. Community Information Resource Centres (CIRCs): Over 2,400 last-mile centres across India delivering digital and emerging-tech literacy, demystifying AI for rural, tribal, and marginalized communities, and enabling access to digital entitlements. 2. Soochnapreneurs (Information Entrepreneurs): A phygital, hyperlocal digital entrepreneurship model combining physical community presence with digital tools, trained rural women and youth who serve as last-mile intermediaries bridging communities with digital entitlements, government schemes, and rights-based services. They document algorithmic exclusion in welfare delivery and act as frontline interpreters of AI and automated systems, ensuring no one is left behind by digital-first governance. 3. Samarth Soochnapreneur Program: A disability-inclusive extension training persons with disabilities as information entrepreneurs, advancing accessibility, dignity, livelihoods, and meaningful inclusion in the digital and AI age. 4, Internet Roshni: Bringing connectivity, digital literacy, and rights-awareness to tea garden workers and their families in Assam and other plantation belts-communities historically excluded from digital and AI ecosystems. 5. Digital Clusters for Artisan Communities: Empowering weavers, craftspeople, and traditional artisans with digital tools, e-commerce access, and cultural-heritage documentation-safeguarding indigenous knowledge from extractive AI practices and enabling fair market participation. 6. Digital Didis: Grassroots women cadres advancing gender-inclusive digital literacy and supporting redress for AI-driven harms in welfare, identity, credit, and online safety. 7. Wireless for Communities (W4C): DEF's community-owned connectivity initiative ensuring meaningful access as a precondition for AI inclusion in unserved and underserved geographies. 8. Digital Citizen Summit (DCS): DEF's flagship Global South-led international multistakeholder platform convening civil society, government, technologists, and grassroots voices on digital rights, inclusion, and AI governance. These initiatives demonstrate that community-led, rights-based, and inclusive AI governance is both possible and essential. DEF stands ready to contribute its grassroots expertise to the Global Dialogue on AI Governance.