Global Digital Justice Forum (GDJF)

We believe that the Global Dialogue would be successful if outcomes reflect a consensus on the following: Cluster 1 Recognizes the Common But Differentiated Responsibilities principle as a ground norm for international AI governance – based on international solidarity and the need to redress the long term consequences of extractivism in the digital economy disproportionately borne by the South. Underscores the need to protect local, traditional, and indigenous knowledge in the AI transition, particularly the right of all peoples to technological autonomy. Cluster 2. Affirms the capability to expand life choices for all and foster critical thinking as foundational for all societies. Recommends establishing an international public financing mechanism to build AI capabilities in the Global South, through a 'CERN for AI' – a distributed network of AI research centers, coordinated by a central hub. Recognizes that innovation choices must remain rooted in the full sovereignty of all nations and peoples. Cluster 3. Eschews voluntary safety commitments and acknowledges the need for binding public accountability standards in the development and deployment of AI. Adopts the 'precautionary approach' as the baseline for AI development and deployment. Calls for a gender equality perspective in AI innovation and pledges zero tolerance for sexual/ gender-based violence resulting from AI systems. Cluster 4. Underscores the need for states to respect, protect and promote human rights, including due diligence, mitigation and remedy of human rights violations in digital value chains. Endorses the call from OHCHR for immediate moratoria of high-risk AI systems and the prohibition of applications inherently incompatible with international human rights law. Emphasizes the need for post-deplyment monitoring of AI through established human rights mechanisms (Human Rights Council, UN Treaty Bodies) Overall. Makes recommendations on how the WSIS and GDC implementation architecture can be effectively utilised to further actions along each of the thematic clusters.

• Safe, secure and trustworthy AI
• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights

1. Safe, secure and trustworthy AI The International AI Safety Report 2026(https://tinyurl.com/4dw27f9v) finds rapidly growing evidence of AI harms. Current approaches relying on voluntary safety commitments have failed. An urgent shift to a 'precautionary approach' in AI development and deployment is needed, alongside safeguards for societal participation towards a just AI transition. This requires an independent mechanism for: (a) mandatory ex-ante risk assessments for high-risk AI systems, and (b) corporate and public accountability across the AI supply chain through post-deployment audits. 2. AI capacity-building In July 2025, the UN Secretary General called for urgent financing to build core AI capacities in developing countries. The private sector avoids high-risk, low-return projects-exactly what such infrastructure requires(https://tinyurl.com/365cd584). Meanwhile, AI-for-social-good initiatives often fail to address local needs or foster human freedoms and digital sovereignty(https://tinyurl.com/3dbjced8). Thus, innovative international public financing mechanisms are critically needed for digital, data, and AI infrastructure in the Global South. 3. Social and other implications of AI As the UN Secretary General has highlighted, AI is not just a technological trend but a force of systemic transformation. Dominant AI trajectories are accelerating socio-economic inequalities, decimating social memory, undermining gender equality, and reinforcing infrastructural dependencies. We must examine AI from a development perspective-attending to its social, economic, cultural, and environmental aspects(https://tinyurl.com/2sj26cru). In the age of LLMs, the implications for linguistic diversity and authenticity must be addressed to challenge knowledge colonialism. 4. Human rights The UN GDC commitment to uphold international human rights law throughout the AI lifecycle must be upheld. This requires mechanisms guaranteeing peoples' right to choose technologies that best suit them. It also entails implementing UN Human Rights recommendations (A/HRC/48/31), including moratoria on procuring and using high-risk AI systems and prohibiting applications incompatible with international human rights law until adequate safeguards, oversight, and accountability are in place.

AI and sustainable development: The Dialogue must explore pathways to address the ecological, labour and knowledge extraction at the heart of the current AI paradigm. AI innovation needs urgent course-correction - a reckoning with its human and ecological costs. International economic law - in trade, taxation and IP - thwarts development autonomy in the current AI paradigm, and must be redesigned for global justice. AI governance depends entirely on and cannot be disembedded from global governance for structural justice. AI and gender equality: Gender-based exclusion, violence and exploitation are an integral feature of the dominant AI paradigm. The proliferation of AI-generated deepfakes, including non-consensual intimate imagery, has intensified gender-based violence, disproportionately targeting women and gender-diverse people. Algorithmic architectures of social media reinforce gender conservative performance, silencing those who do not conform. AI value chains are constitutive of women's precarious and invisible labour in content moderation, data labelling, and more. A resource-intensive AI paradigm intensifies the crisis of natural resources that sustain women's reproductive labour. An intersectional lens is needed to effectively respond to the gendered shocks inherent in the AI paradigm. AI commons: The focus on open-source software, open source computing, open data and open artificial intelligence models are necessary, but not sufficient, for democratising the innovation dividends of the AI innovation paradigm. The Dialogue should explicitly examine the enabling conditions, such as public incentives, social licensing frameworks for data and AI innovation, and community-led data stewardship models, that would make a productive dynamic between digital public infrastructure and AI commons initiatives possible at all levels of the AI stack.

The Global Digital Justice Forum is a civil society coalition committed to furthering digital justice for the people of the Global South. We see the following challenges that may impede the Dialogue's potential to meaningfully address the AI governance deficit : - Debates on safety and risks in AI tend to be dominated by a pro-industry approach. They are not linked to binding global baselines for technology assessment grounded in international human rights law. - With immense pressure on developing countries to catch up in the AI economy, they are arm-twisted by powerful countries into extractive digital trade and cooperation arrangements. These are ill-suited to their own context and not accountable to their citizens. IP frameworks have scooped up local knowledge without compensation to producers and creators and the lack of an international digital tax regime and the inordinate power of digital TNCs has affected developing countries adversely. - The crisis in international public financing for development, and a global impasse in addressing this, makes it difficult to mobilise resources for AI capabilities in the Global South. - AI is often viewed in a "means of implementation" frame rather than a constitutive force of structural transformation – this results in design tweaks rather than systemic solutions to data extractivism, technological colonialism and environmental destruction. - The narrative power of Big AI tends to crowd out local AI initiatives rooted in frugal AI principles. If the Global Dialogue puts people and public interest at the centre and mobilizes moral leadership to steer a new grammar of global AI governance grounded in human rights, sustainable development, and structural justice, we will have made a breakthrough. There is an opportunity here to recognize and encourage AI models that are fair and frugal – grounded in people's wellbeing, attentive to ecological impacts and shaped by participatory methods.

The Global Dialogue on AI can play the following role in advancing international cooperation in AI governance: (1) Protection and promotion of human rights. The Dialogue must advance international cooperation for encoding public accountability and corporate liability frameworks in complex, multi-actor AI value chains, in compliance with international human rights law. (2) Building equitable outcomes for development in the AI society. The Dialogue must effectively catalyze international cooperation among countries as well as communities of practice for democratizing the benefits of AI. Further, it must center the standpoints of the South in the implementation of the UN HLAB-AI's proposals for i) a global capacity-building network on AI and ii) international standards exchange. Equity also implies forays into a new AI paradigm that is non-extractive, and built for sustainability and social care – taking into account diverse socio-cultural and economic contexts. (3) Integrated vision of data and AI governance. AI governance should be based on data justice – cognizant of the potential harms of AI to people and the planet and the necessary role of collective stewardship of data resources for equitable and democratic AI innovation.

1. Linkages with WSIS and GDC implementation architecture. The Global Dialogue should build appropriate linkages with (a) The UN CSTD Working Group on Data Governance, (b) the International Scientific Panel on AI and (c) the WSIS+20 Internal Task Force on Financing Mechanisms 2. Frameworks that the Global Dialogue should build upon. In addition to the international human rights framework, the Dialogue must build upon the conceptual rubrics of: - the Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (Council of Europe) - the UNESCO Recommendation on the Ethics of AI. - the BRICS Leaders' Statement on the Global Governance of AI - especially, its focus on development sovereignty and the socio-economic and cultural impacts of the AI revolution.

As civil society groups have submitted in their open letter to the co-chairs(https://tinyurl.com/3d7enwt7), it is critical that the structure/format of the Global Dialogue create the enabling conditions for meaningful participation. The hegemony of certain powerful voices should not drown out alternative perspectives. Meaningful participation is a human right in itself, and, as called for by the international human rights framework, it must be grounded in the principles of non-discrimination, equity, and justice, while giving special attention to marginalized people and groups (e.g. based on gender, caste, class, religion, ethnicity, age etc.) Also, an appropriate structure and creative formats may only be a starting point for inclusion. Meaningful participation also hinges on a range of other conditions: financial and logistical support for underrepresented communities, particularly from the Global Majority, for in-person participation in the Global Dialogue (see e.g. IGF Trust Fund - https://www.intgovforum.org/en/content/igf-trust-fund). adequate space for civil society perspectives, especially from marginalized communities, during discussions and in the preparation of the Summary document. open and transparent consultations with civil society at the Dialogue, to gather insights and perspectives that may otherwise be drowned out in broader multistakeholder spaces. The UNESCO Recommendation on Open Science highlights that beyond the scientific and technical community, the meaningful engagement of civil society is essential in the pursuit of science. Special attention must be given to ensuring the effective participation of organizations representing marginalized people and groups, such as ethnic groups, Indigenous Peoples, peasants and other small-scale food providers, workers, women and girls, LGBTIQ+, youth, etc. Finally, it is important to appoint a civil society representative who will serve as a liaison with the Co-Chairs and provide inputs to the drafting of the Summary document of the Global Dialogue.

Voices/perspectives that are underrepresented in AI governance discussions: - Communities adversely affected by extractivism in the production chains of AI, particularly those in the majority world. - People's movements that have called out the harmful impacts of digitalization, datafication and Big Tech control over socio-economic and cultural domains – including, food sovereignty, ecological justice, trade justice, labour rights, public health, gender equality, cultural rights, Indigenous Peoples' rights, and the right to education and knowledge. - Technologists committed to commons-based innovation; decentralised platform, data, and AI models; public technology standards; and ethical design, including feminist data science approaches. How can these voices/perspectives be included: - Enhancing the space for civil society participation through the suggestions discussed in our answer to Q15, and putting in place mechanisms to address power asymmetries. - Ensuring that in the Summary/ Outcome Document, convergences and divergences are recorded under each thematic cluster in order to prevent diverse perspectives that challenge the dominant discourse from being silenced/invisibilised.

- Adopt session formats that allow civil society and other non-governmental and governmental stakeholders to respond to each other in real time. There should be a minimum of 50% rotating time allocated for non-governmental stakeholders to engage in dialogue with each other and with governments. - Institute a Multistakeholder Advisory Group (MAG) process for all stakeholder constituencies to weigh in on the key outcomes. This is like the Internet Governance Forum (IGF) and Informal Multistakeholder Sounding Board in the WSIS+20 review. - Focus on both a) overarching, transversal discussions that address cross-cutting AI governance issues, including normative frameworks, human rights standards, ethical principles, and global regulatory coherence and b) sector-specific deep dives that examine AI's impact, risks, and opportunities in key areas such as healthcare, education, food and nutrition, labor, finance, climate action and biodiversity protection, gender-based violence, corporate accountability, and security.

As the UNDP's Human Development Report 2025 underscores, the transformative impacts of AI arise not merely from technical affordances but from the dynamic ways these affordances interact with social forces. The core challenge in the global governance of AI is to ensure that our efforts to 'bridge the gendered, racialised, classed AI divides' do not result in adverse incorporation into the unequal AI paradigm. We need a systemic transformation that leads to substantive capabilities and freedoms for all people. The core policy touchstone in this regard is Agenda 2030 in which states reasserted their commitments as follows: "We reaffirm that every State has, and shall freely exercise, full permanent sovereignty over all its wealth, natural resources and economic activity. We will implement the Agenda for the full benefit of all, for today's generation and for future generations. In doing so, we reaffirm our commitment to international law and emphasize that the Agenda is to be implemented in a manner that is consistent with the rights and obligations of States under international law. We reaffirm the importance of the Universal Declaration of Human Rights, as well as other international instruments relating to human rights and international law." The Aarhus Convention's provisions for proactive public participation, access to information, and access to justice provide a highly relevant framework to imagine societal rights to participation in shaping the future trajectories of the AI revolution on environmental and other matters. We would also like to point to the collective civil society wisdom from UNESCO's Global CSO Network on AI Ethics as an important reference point. One of the reports of the network(https://tinyurl.com/mryjek3s) provides a useful mapping of the baselines for human rights and global justice in AI value chains.