Global Innovation Square: A Raisina-ORF Initiative

The first Global Dialogue on AI Governance would be a success if it helps shift the global conversation from AI governance as rule-setting alone to AI governance as an enabler of responsible, inclusive, and replicable pathways for AI diffusion that serve people and development. A successful Dialogue should produce three broad outcomes. First, it should create a shared understanding of how AI can be deployed for public good across diverse development contexts, particularly in areas such as health, education, agriculture, climate resilience, social protection, and public service delivery. Second, it should identify practical pathways through which AI capabilities can diffuse more equitably, including access to compute, quality datasets, open and interoperable models, skills, institutional capacity, and financing. Third, it should enable countries and stakeholders to exchange replicable approaches, including policy frameworks, public digital infrastructure, capacity-building models, and governance practices that can be adapted to national contexts. In this sense, the first Dialogue should lay the foundation for an ongoing global conversation on AI diffusion, anchored in the principle that AI governance must protect people, empower societies, and support inclusive progress.

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

First, AI capacity-building is essential to ensure that countries are not only consumers of AI systems, but active participants in shaping, adapting, and governing them. This includes building human capital, institutional capacity, technical expertise, public-sector readiness, and the ability to deploy AI responsibly across development sectors. Second, the social, economic, ethical, cultural, linguistic, and technical implications of AI must be addressed together. AI diffusion will affect labour markets, public services, education, health, culture, language diversity, and social inclusion. Governance discussions must therefore move beyond technical risk alone and consider how AI systems interact with local contexts, development priorities, and existing inequalities. Third, open-source software, open data, and open AI models are important enablers of access, innovation, and replicability. They can lower barriers to entry, support local experimentation, and allow countries and communities to adapt AI applications to their own needs. At the same time, openness must be accompanied by appropriate safeguards, documentation, accountability, and capacity to use these resources meaningfully and responsibly. While safe, secure, and trustworthy AI remains central to global AI governance, the Independent International Scientific Panel on AI is specifically mandated to provide annual evidence-based scientific assessments of AI opportunities, risks, and impacts. The Global Dialogue can complement this work by focusing on implementation-oriented priorities, especially capacity-building, responsible AI diffusion, and replicable approaches for AI for people and development.

Yes. A key cross-cutting issue is responsible and inclusive AI diffusion. Current AI governance debates often focus on principles, risks, and regulation, but less on how AI capabilities actually spread across countries, sectors, institutions, and communities. For many developing countries, the core question is not only how AI should be governed, but how access to AI infrastructure, skills, datasets, models, financing, and institutional capacity can be expanded responsibly and equitably. A second issue is replicability. The AI Dialogue should examine which governance and deployment models can be adapted across different national contexts, especially in the Global South. This requires attention to local languages, data availability, public-sector capacity, digital public infrastructure, affordability, and institutional readiness. Finally, the Dialogue should strengthen the global conversation on AI for people and public purpose, focusing on how AI can improve outcomes in health, education, agriculture, climate resilience, social protection, and public service delivery.

India's experience shows how AI capacity-building, open AI resources, and responsible AI diffusion are closely linked. Through the IndiaAI Mission, India is advancing several relevant priorities, including compute access, indigenous Large Multimodal Models and domain-specific foundational models, AIKosh as a unified platform for datasets, models and AI sandboxes, application development for socio-economic transformation, startup financing, FutureSkills and Data and AI Labs in Tier 2 and Tier 3 cities, and safe and trusted AI. These efforts create opportunities to democratise AI access and scale AI for people in areas such as health, education, agriculture, climate resilience, and public service delivery. However, challenges remain around equitable access, institutional capacity, local-language AI, responsible use of open models, and safeguards for trusted deployment. The AI Dialogue can help countries exchange such experiences and identify replicable pathways for responsible AI diffusion.

The AI Dialogue can play an important role in advancing international cooperation by serving as a platform for implementation-oriented and inclusive conversations on AI governance. Its value should lie in connecting global principles with practical pathways for responsible AI diffusion, especially for countries and communities that currently face capacity, infrastructure, data, and skills gaps. The Dialogue can help countries share experiences on what works in building AI ecosystems, including compute access, open datasets, open models, skilling initiatives, public-sector adoption, startup support, and safeguards for trusted deployment. It can also support the exchange of replicable models that can be adapted to different national contexts, rather than promoting one-size-fits-all approaches. Importantly, the Dialogue can help broaden international cooperation beyond risk management alone. It can create space for a global conversation on AI for people, focused on how AI can support health, education, agriculture, climate resilience, social protection, and public service delivery. By bringing together governments, international organisations, industry, academia, civil society, and technical communities, the Dialogue can strengthen trust, reduce fragmentation, and identify practical areas for cooperation. In this way, the AI Dialogue can act as a bridge between global norm-setting and local implementation, ensuring that AI governance supports inclusion, development, and shared public value.

The AI Dialogue should build upon and connect with the Global AI Impact Commons, which offers a practical model for advancing AI cooperation around real-world impact. The Commons is designed as a global platform to enable the replication of proven AI solutions across sectors and regions, with documented impact stories from over 30 countries and more than five sectors. It is also an outcome of the India AI Impact Summit 2026 and the Working Group on AI for Economic Growth and Social Good, led by India, the Netherlands, and Indonesia. The added value of the AI Dialogue would be to provide a broader UN-led platform to take such efforts from documentation to diffusion. It can help countries and stakeholders learn from impact stories, identify what makes AI solutions replicable, and support their adaptation across different national and institutional contexts. The Dialogue can also connect evidence of AI impact with policy discussions on capacity-building, open datasets, open models, compute access, skills, safeguards, and financing. By engaging with the Global AI Impact Commons, the AI Dialogue can help shift global AI governance from abstract principles alone to practical pathways for AI for people, enabling responsible, inclusive, and measurable AI diffusion for development.

The AI Dialogue should be structured as a continuous process, not a one-off event. It should include more frequent informal stakeholder consultations, both virtual and in person, along with written inputs from diverse stakeholders. To strengthen transparency, summaries or minutes of consultations could be shared after meetings, where appropriate, along with public outcome notes or press releases. The format could combine plenaries, thematic roundtables, regional consultations, expert briefings, and solution-focused sessions. This would make the Dialogue more participatory, transparent, and implementation-oriented. Additionally, workshops focused on use cases and the exchange of knowledge / experiences should be curated across the world to support AI diffusion and the replicability of solutions

Voices from the Global South remain underrepresented in global AI governance discussions, particularly those of developing-country governments, local innovators, civil society, researchers, and communities directly affected by AI deployment such as those whose impact stories are documented by the Global AI Impact Commons. They can be included through targeted regional consultations, travel support, multilingual engagement, written submissions, and dedicated sessions on development priorities, AI diffusion, and replicable models for AI for people.

The format could combine plenaries, thematic roundtables, regional consultations, expert briefings, and solution-focused sessions. This would make the Dialogue more participatory, transparent, and implementation-oriented.

IndiaAI Mission: The IndiaAI Mission offers an ecosystem-based approach to AI capacity-building and responsible deployment. It brings together compute access, indigenous AI models, AIKosh for datasets and sandboxes, application development, startup financing, FutureSkills, Data and AI Labs in Tier 2 and Tier 3 cities, and safe and trusted AI. Global AI Impact Commons: The Global AI Impact Commons offers a platform-based approach to documenting and sharing real-world AI use cases. It can help countries study, adapt, and replicate practical AI solutions across sectors and contexts, supporting responsible AI diffusion for people and development.