CCAOI

Success for the Global Dialogue on AI Governance ( "AI Dialogue") means building actionable consensus across its thematic clusters, reflected substantively in the Co‑Chair Summary. To achieve this, the AI Dialogue should ensure: - Meaningful multistakeholder engagement: Participation must be inclusive, with particular emphasis on the Global South and civil society. The AI Dialogue should operationalise mechanisms that enable diverse voices to contribute effectively, setting a model for other AI governance processes. - Alignment with global processes: Clear pathways must be identified to connect the AI Dialogue with ongoing initiatives such as the AI Summit series, OECD, GPAI, IGF, CSTD, WSIS Forum, and ITU's AI for Good, as well as relevant UN bodies. For example, the Independent International Scientific Panel could present findings to WSIS communities and the Global IGF, and vice versa. - Financing and capacity building: The AI Dialogue should advance proposals from the Global Digital Compact, including the Global Fund on AI, and establish financing mechanisms to strengthen AI capabilities in the Global South. - Accountability architecture: A foundation must be laid for concrete commitments—respecting and promoting human rights—by both member states and the private sector, alongside mechanisms to hold these actors accountable. Additionally, the AI Dialogue should make explicit recommendations on how existing WSIS and GDC implementation structures—including the UN CSTD Working Group on Data Governance, the International Scientific Panel on AI, and the WSIS+20 Internal Task Force on Financing Mechanisms—can be mobilised for follow‑through. By embedding inclusivity, alignment, financing, and accountability, the Dialogue can deliver outcomes that are both substantive and actionable, shaping a credible and responsive global AI governance framework.

• 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 There is an increase in real-world AI risks-malicious use, malfunctions, and systemic harms-is mounting, while voluntary safety pledges remain insufficient. The Dialogue should build on the Independent Scientific Panel's work to establish a clear taxonomy of risks, particularly those affecting vulnerable populations across privacy, discrimination, resource access, cybersecurity, and democracy. It should advance a precautionary approach as a binding standard and develop mechanisms for safe deployment in the Global South, including incident reporting, risk assessment frameworks, and data auditing. 2. AI Capacity-Building The Dialogue should secure concrete commitments from governments and the private sector, while exploring innovative pathways for international public financing mechanisms to strengthen AI capabilities in the Global South. 3. Social, Economic, Ethical, Cultural, Linguistic, and Technical Implications AI is reshaping societies far beyond technology. Current trajectories risk deepening inequality, eroding cultural commons, and entrenching dependency. The Dialogue must adopt a human development lens and be attentive to social, economic, cultural, and environmental dimensions-and address knowledge colonialism, particularly in the context of large language models. 4. Human Rights The AI Dialogue must operationalise the commitment in paragraph 22 of the UN Global Digital Compact to uphold international human rights law throughout the lifecycle of AI technologies, ensuring that rights are protected and promoted in both principle and practice.

AI and Sustainable Development The AI Dialogue must address environmental and labour concerns, exploring how innovation can be refocused toward sustainable development goals. Embedding sustainability into AI design and deployment is essential to ensure technology supports inclusive growth rather than exacerbating inequities. AI and Gender Equality Gender-based exclusion and exploitation remain integral to dominant AI paradigms. Governance discussions must adopt an intersectional lens, ensuring that gender equality is mainstreamed across all aspects of AI policy, design, and oversight. AI Commons Open-source software, open data, and open AI models are critical to democratizing innovation. Equally important is building technological and institutional support for productive synergies between digital public infrastructure and community-led digital commons. Enabling conditions-such as public incentives, social licensing frameworks, and diverse models of data stewardship-can foster equitable participation across the AI stack. AI and Privacy In many Global South countries, existing data protection frameworks inadequately address AI-driven data aggregation. For example, India's Digital Personal Data Protection Act emphasizes consent-based processing, yet AI systems rely on large-scale data collection, reuse, and inference that extend beyond traditional consent boundaries. This creates accountability gaps around dataset provenance, auditability, and model deployment. Compliance with privacy norms alone cannot mitigate risks of bias, opacity, or re-identification. AI-specific governance frameworks are needed to ensure responsible data use and safeguard rights.

Safe, Secure and Trustworthy AI AI systems often embed and perpetuate biases due to the underrepresentation of many communities—particularly from the Global South—in global networks, processes, research, and datasets. Data is frequently extracted from these regions without equitable benefit sharing, while corporations overlook local contexts in technological development and deployment. This results in inequitable outcomes, heightened risks, security vulnerabilities, and diminished trust. Addressing these challenges requires risk management frameworks grounded in human rights, adaptable to diverse contexts, and supported by locally‑rooted research, capacity building, and accountability mechanisms. Crucially, affected communities must be meaningfully included in shaping and participating in AI governance. Debates on safe and trustworthy AI and human rights are currently dominated by voluntary, industry‑led safety commitments. These lack binding global baselines for technology assessment aligned with international human rights law. For developing countries, immense pressure to rapidly build domestic AI capacities makes adopting a precautionary approach to AI development and deployment particularly challenging. AI Capacity‑Building Equitable distribution of AI resources and benefits requires investment in financing, knowledge transfer, and capacity building, alongside support for research and infrastructure in the Global South. Protection and Promotion of Human Rights Many AI technologies are being deployed without human rights impact assessments. In the Global South, governance processes often outpace local capacity to engage, limiting meaningful participation and the ability to surface human rights concerns. This risks reproducing existing inequalities through governance frameworks and undermines the rights of communities. Embedding human rights safeguards across the AI lifecycle is essential to ensure inclusive, equitable, and accountable AI governance.

The AI Dialogue on AI can serve as a vital platform for advancing international cooperation in AI governance by ensuring meaningful participation across all stakeholder communities and creating a level playing field for the Global South. While doing that the AI Dialogue must prioritize the protection and promotion of human rights, embedding accountability and liability frameworks across complex AI value chains in line with international law. Governments, industry, and civil society should collaborate to ensure that AI systems respect rights and uphold public trust. The AI Dialogue should focus on capacity‑building, catalyzing triangular and South‑South cooperation to direct AI development toward sustainable development and global peace. This includes operationalizing the global capacity‑building network on AI and advancing international standards exchange, as recommended by the UN High‑Level Advisory Body on AI. The AI Dialogue must promote an integrated vision of data and AI governance, attentive to intersections between stewardship of the social data commons, elimination of algorithmic harms, and democratization of innovation dividends. By embedding inclusivity, capacity‑building, and accountability, the AI Dialogue can strengthen coherence across global processes and ensure AI governance advances both human rights and sustainable development.

The AI Dialogue should not operate in isolation but build strong linkages with existing global regional mechanisms and Frameworks. First, it must connect with the Internet Governance Forum (IGF), leveraging its established multistakeholder processes and National/Regional Initiatives (NRIs) to ensure inclusivity and continuity. Equally important are linkages with the WSIS and Global Digital Compact (GDC) implementation architecture, including: - The UN CSTD Working Group on Data Governance, to align AI governance with broader data frameworks. - The International Scientific Panel on AI, to integrate evidence‑based insights into deliberations. - The WSIS+20 Internal Task Force on Financing Mechanisms, to advance sustainable financing for AI capacity‑building, particularly in the Global South. The AI Dialogue should also build upon conceptual frameworks that embed human rights and ethical principles. These include the Council of Europe's Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law, and the UNESCO Recommendation on the Ethics of AI, both of which provide normative rubrics for safe, secure, and trustworthy AI. From a development perspective, the AI Dialogue must integrate outcomes from other initiatives to ensure AI governance reflects diverse cultural and developmental contexts. The added value of the AI Dialogue lies in its ability to connect these initiatives coherently, avoid duplication, and create actionable pathways that bridge normative frameworks, scientific evidence, financing mechanisms, and grassroots participation. By embedding inclusivity and accountability, the Dialogue can strengthen legitimacy and provide a responsive architecture for global AI governance—anchored in human rights, sustainable development, and equitable participation.

Governments should bring regulatory perspectives and public policy priorities, particularly from the Global South, highlighting developmental concerns, digital divides, and capacity constraints. The private sector can contribute technical expertise, best practices, and risk assessment mechanisms, while committing to transparency and accountability. Civil society and academia can contribute to creating a bottom‑up agenda, representing affected communities, interrogating power asymmetries, and grounding deliberations in human rights and established principles. Recommended Formats and Structures: - Multistakeholder Advisory Group/Steering Committee: To guide the AI Dialogue, ensure representation, and provide expertise and resources. - Balanced Thematic Multistakeholder Working Groups: To support Co‑Chairs on key themes with diverse stakeholder input. - Regional Consultations: Leveraging the IGF Ecosystem such as the NRIs, Dynamic Coalitions, Best Practice Forums, Policy Networks, and youth initiatives to surface concerns and policy issues across tiers. Key Practices: - Continuous, robust consultation: Regular open consultations, with mechanisms such as zero drafts circulated for stakeholder comments, as used in WSIS+20. - Shared spaces and mingled tracks: Bringing governments, industry, civil society, technical community and academia together to deliberate and co‑create integrated outcomes. - Transparent modalities: Public summaries explaining how stakeholder input shaped final decisions and outcomes. To ensure meaningful participation, governments and non‑government actors must provide financial and logistical support for underrepresented communities, particularly from the Global South, to join in person. Within the Dialogue, civil society perspectives should be given dedicated space, including exclusive consultations and a civil society liaison to support Co‑Chairs in drafting the Summary Outcome. Special attention must be paid to marginalized groups—indigenous peoples, workers, women, LGBTIQ+, youth, and small‑scale businesses, so their voices inform both process and output. This inclusive structure will strengthen legitimacy, prevent discursive hegemony, and ensure the AI Dialogue reflects diverse realities in shaping global AI governance.

Apart from ensuring broad multistakeholder participation, the AI Dialogue must meaningfully engage voices from the Global South, particularly civil society, communities advancing gender equality, labour rights, indigenous peoples, as well as social scientists, ecologists, environmentalists, and youth. Their perspectives are essential not only for inclusivity but also for shaping outcomes that reflect diverse realities. Crucially, these groups should be involved in agenda‑setting, decision‑making, and framing of outputs, rather than being confined to consultation alone. To embed their voices, the AI Dialogue should: - Enhance civil society participation spaces, ensuring equitable access to discussions and decision‑making platforms. - Institutionalize transparency by recording both convergences and divergences under each thematic cluster in the Summary/Outcome Document. This prevents non‑hegemonic voices from being drowned out and ensures accountability to diverse perspectives. - Create mechanisms for feedback loops, where marginalized groups can see how their input has shaped deliberations and final recommendations. - Support capacity building and accessibility measures, including language diversity, digital inclusion, and resources for participation, so Global South actors can engage on equal footing. By embedding these practices, the AI Dialogue will not only broaden participation but also strengthen legitimacy. Meaningful engagement of underrepresented voices ensures that governance frameworks are responsive to global realities, uphold human rights, and advance sustainable development. In doing so, the Dialogue can become a model for inclusive, participatory AI governance—where diverse communities shape the digital future together.

To foster meaningful and dynamic engagement, the AI Dialogue should adopt multistakeholder participatory and deliberative formats that encourage diverse perspectives and enable collaborative decision‑making. Transparency is essential: a clear process roadmap, calendar of consultations, public repository of written submissions, and summaries explaining how stakeholder input shaped outcomes will build trust and legitimacy. Session formats must allow stakeholders—especially civil society, non‑governmental actors, and governments—to respond to each other in real time, ensuring genuine dialogue rather than parallel monologues. Establishing a Multistakeholder Advisory Group (MAG), modelled on the IGF or the Informal Multistakeholder Sounding Board of the WSIS+20 review, would provide a structured mechanism for all constituencies to weigh in on key outcomes before they feed into the final multilateral adoption process. The AI Dialogue should balance transversal discussions on cross‑cutting AI governance issues—normative frameworks, human rights standards, ethical principles, and global regulatory coherence—with sector‑specific deep dives. These deep dives could explore AI's impact, risks, and opportunities in areas such as healthcare, education, food and nutrition, labour, finance, climate action and biodiversity protection, gender‑based violence, corporate accountability, and security. By combining inclusive formats, transparent processes, and thematic breadth, the Dialogue can create a participatory architecture that not only reflects diverse voices but also strengthens legitimacy and responsiveness. This approach will ensure that outcomes are both substantive and actionable, positioning the Dialogue as a credible model for global AI governance.

A central challenge of AI governance lies in bridging the trust and safety deficit and addressing the AI divides-whether gender, technological, or socio-economic-that risk creating an unequal AI paradigm. To overcome these challenges, governance must be anchored in inclusive, transparent, and meaningful processes and outcomes. The AI Dialogue can draw valuable lessons from the history of multistakeholder Internet governance. The World Summit on the Information Society (WSIS) and its Informal Multistakeholder Sounding Board (IMSB) demonstrated how diverse voices can be integrated into global processes. The Internet Governance Forum (IGF), through its Multistakeholder Advisory Group (MAG), has institutionalized representation and inclusivity, while the NETmundial process and the São Paulo Multistakeholder Guidelines provide tested approaches for embedding openness, accountability, and legitimacy in decision-making. These precedents highlight practices that the AI Dialogue should adopt: - Multistakeholder advisory structures to ensure balanced input across constituencies. - Transparent documentation of convergences and divergences in outcomes, preventing dominant voices from drowning out alternative perspectives. - Inclusive consultation mechanisms, ensuring civil society, youth, indigenous communities, and marginalized groups have equitable space in deliberations. - Capacity-building support, particularly for Global South stakeholders, to enable meaningful participation.