Amity Law School, Amity University Noida

A successful first Global Dialogue on AI Governance would produce actionable, inclusive, and interoperable outcomes rather than purely declaratory principles. First, the Dialogue should establish a baseline set of globally agreed principles for safe, secure, and trustworthy AI that are sufficiently flexible to accommodate diverse national regulatory approaches while ensuring minimum safeguards. Second, it should initiate a structured coordination mechanism under the United Nations to promote interoperability across governance frameworks, reducing fragmentation and enabling cross-border collaboration. Third, a key outcome should be the launch of concrete capacity-building initiatives, particularly for developing countries, focusing on access to high-performance computing, datasets, and technical expertise. Without such measures, existing digital divides risk deepening into structural AI inequalities. Fourth, the Dialogue should advance practical guidance on transparency, accountability, and human oversight, including pathways to reconcile these requirements with legitimate intellectual property protections. Fifth, the process should catalyse the development of open and inclusive AI ecosystems, including support for open-source models and public-interest data governance frameworks. Finally, success would be reflected in sustained multi-stakeholder engagement, ensuring that academia, industry, civil society, and governments, especially from the Global South, are meaningfully represented in ongoing governance processes.

Safe, secure and trustworthy AI;AI capacity-building;Interoperability of governance approaches;Transparency, accountability, and human oversight

These priorities reflect the need for a balanced and integrated AI governance framework. Ensuring safe, secure, and trustworthy AI is foundational to public trust and long-term sustainability of AI systems. However, safety must be complemented by transparency, accountability, and human oversight, particularly in high-impact applications, to safeguard fundamental rights and prevent opaque decision-making. Interoperability of governance approaches is critical in the current fragmented regulatory landscape. Divergent national frameworks risk creating compliance burdens, regulatory arbitrage, and barriers to innovation. Developing baseline standards that enable coordination across jurisdictions is therefore essential. At the same time, AI capacity-building must be prioritized to address structural inequalities between developed and developing countries. Access to computational infrastructure, datasets, and technical expertise remains highly uneven. Without targeted interventions, AI governance risks reinforcing global asymmetries in economic and technological power. From an intellectual property perspective, these priorities are deeply interconnected. For instance, transparency obligations must be balanced with trade secret protections, while capacity-building efforts require rethinking access to proprietary AI systems and datasets. Together, these areas enable a governance approach that is not only technically robust and legally coherent, but also equitable and globally inclusive.

A key cross-cutting issue insufficiently captured in the listed themes is the intersection of artificial intelligence governance with intellectual property (IP) regimes. IP frameworks play a central role in shaping: 1) access to AI technologies 2) control over datasets and models 3) distribution of economic benefits Firstly, current uncertainties regarding authorship, inventorship, and ownership of AI-generated outputs create regulatory ambiguity, while strong proprietary controls over data and models may limit transparency, accountability, and equitable access. A second emerging issue is data governance as a strategic public good. The availability, quality, and governance of datasets are foundational to AI development, yet global asymmetries persist in data access and control. There is a need to develop frameworks that promote data-sharing, data sovereignty, and public-interest data commons, particularly for developing countries. Third, the political economy of AI including market concentration and the dominance of a small number of technology actors requires greater attention. Governance discussions must address not only technical risks but also structural power imbalances in AI ecosystems. Finally, the interaction between AI governance and development policy should be foregrounded. AI should be embedded within broader sustainable development objectives to ensure that its benefits are widely distributed. Addressing these cross-cutting issues will be essential for building a coherent, equitable, and future-ready global AI governance framework.

In the Indian and broader Global South context, governance gaps in AI are producing a mixed landscape of accelerated adoption alongside structural vulnerabilities. A primary challenge lies in the absence of interoperable governance frameworks. Divergent regulatory approaches across jurisdictions create uncertainty for cross-border data flows, AI deployment, and compliance, affecting sectors such as fintech, healthcare, and digital public infrastructure. This fragmentation also complicates alignment with emerging global standards. Second, limited AI capacity and infrastructure—including constrained access to high-performance computing, quality datasets, and advanced research ecosystems—restricts domestic innovation. This leads to dependence on proprietary AI systems, raising concerns around technological sovereignty and long-term competitiveness. Third, gaps in transparency, accountability, and human oversight are increasingly visible in high-impact use cases such as automated decision-making in finance, recruitment, and governance. The absence of clear auditability standards and explainability requirements risks undermining public trust and may expose individuals to opaque or biased outcomes. At the same time, intellectual property and data governance tensions affect both access and accountability. Strong proprietary protections over models and datasets can limit transparency and local adaptation, while regulatory ambiguity around AI-generated outputs creates uncertainty for innovators. However, these challenges are accompanied by significant opportunities. India's expanding digital ecosystem and public digital infrastructure provide a foundation for scalable, inclusive AI applications. There is also growing potential to shape globally relevant governance models that balance innovation with equity and rights. Strategic investments in capacity-building, combined with coordinated efforts under the United Nations, can enable more inclusive participation in global AI value chains while addressing systemic governance gaps.

The Global AI Dialogue can play a pivotal role as a neutral, multi-stakeholder platform for advancing international cooperation on AI governance under the United Nations. First, it can facilitate the development of baseline global principles for safe, secure, and trustworthy AI. While binding harmonization may not be feasible, convergence around minimum standards can reduce fragmentation and enable interoperability across national regulatory frameworks. Second, the Dialogue can serve as a coordination hub that connects existing initiatives across governments, international organizations, industry, and academia. By mapping ongoing efforts and identifying gaps, it can help avoid duplication and promote coherent global governance architectures. Third, it can advance capacity-building cooperation, particularly for developing countries. Through partnerships, resource mobilization, and knowledge-sharing mechanisms, the Dialogue can support access to computational infrastructure, datasets, and technical expertise, thereby addressing structural inequalities in AI development. Fourth, the Dialogue can provide a space to address cross-border challenges, including data governance, AI safety standards, and accountability mechanisms. Developing shared approaches to these issues is essential for managing risks that transcend national boundaries. Fifth, it can promote inclusive participation in norm-setting, ensuring that perspectives from the Global South, civil society, and technical communities are meaningfully integrated into global governance discussions. Finally, the Dialogue can catalyse the development of practical tools, such as model guidelines, best practices, and voluntary frameworks, which can be adapted by countries based on their regulatory contexts. Through these functions, the AI Dialogue can move global discussions from fragmented debates toward structured, sustained, and cooperative governance processes.

The Global AI Dialogue should build upon and connect with a range of existing international initiatives to avoid fragmentation and leverage established expertise. Key multilateral efforts include work under the United Nations system (such as UNESCO's Recommendation on the Ethics of AI), as well as standard-setting and policy coordination initiatives by the OECD (AI Principles), G20 (AI policy guidance), and the Global Partnership on AI (multi-stakeholder research and best practices). In addition, regional regulatory developments—such as the European Union's AI regulatory framework—are shaping emerging norms on risk-based governance, safety, and accountability. Technical and governance coordination efforts, including standards development through bodies like the International Organization for Standardization and the International Telecommunication Union, also provide important foundations for interoperability and technical alignment. Despite these initiatives, the global AI governance landscape remains fragmented and unevenly inclusive. The added value of the AI Dialogue lies in its ability to act as a convening and integrating platform that: - Bridges policy, technical, and legal discussions across institutions - Enhances interoperability between different governance frameworks - Ensures inclusive participation, particularly from developing countries and underrepresented stakeholders - Connects high-level principles with practical implementation pathways Importantly, the Dialogue can help align these initiatives around shared priorities such as capacity-building, transparency, and equitable access, while identifying gaps that require coordinated global responses. By building on existing mechanisms rather than duplicating them, the AI Dialogue can contribute to a more coherent, collaborative, and effective global AI governance ecosystem.

Effective AI governance requires structured, multi-stakeholder participation, and the AI Dialogue can operationalize this through clearly defined roles and engagement formats under the United Nations. Stakeholder Contributions 1) Member States: Provide regulatory perspectives, share national experiences, and contribute to the development of interoperable governance approaches. 2) Academia and Research Institutions: Offer evidence-based analysis, foresight on emerging risks, and interdisciplinary policy inputs. 3) Private Sector: Contribute technical expertise, operational insights, and best practices on safety, deployment, and innovation. 4) Civil Society: Represent public interest concerns, including human rights, inclusion, and societal impacts. 5) Technical Community and Standards Bodies: Support the development of technical benchmarks, audit mechanisms, and interoperability standards. Recommended Format and Structure 1) Thematic Working Groups Organize the Dialogue around core thematic clusters (e.g., safety, capacity-building, human rights), with each group producing targeted outputs and recommendations. 2) Hybrid Consultation Model Combine: a) open written submissions (broad participation) b) curated expert roundtables (depth and technical rigor) 3) Regional Consultations Ensure geographically balanced participation, particularly from developing countries, to reflect diverse priorities and contexts. 4) Outcome-Oriented Sessions Each engagement phase should aim to produce: a) policy briefs b) model guidelines c) best practice frameworks 5) Iterative and Continuous Process Establish the Dialogue as an ongoing mechanism, with periodic reviews, updates, and follow-up actions rather than a one-time event. 6) Transparency and Accessibility Measures Publish summaries, key findings, and recommendations to ensure accountability and wider dissemination. Such a structured approach would enable the AI Dialogue to move beyond discussion toward coordinated, actionable, and inclusive global governance outcomes.

Global discussions on AI governance continue to reflect asymmetries in participation and influence, with several critical voices underrepresented. First, developing countries and Global South stakeholders remain insufficiently included in norm-setting processes, despite being significantly affected by AI deployment. Limited institutional capacity, resource constraints, and restricted access to technical infrastructure often hinder meaningful participation. Second, non-English-speaking and culturally diverse communities are underrepresented, leading to governance frameworks that may not adequately reflect linguistic diversity or local socio-cultural contexts. Third, civil society organizations, grassroots groups, and affected communities—including workers impacted by automation, marginalized populations, and vulnerable groups—often lack direct representation in technical and policy discussions. Fourth, interdisciplinary academic perspectives, particularly from law, social sciences, and humanities, are sometimes overshadowed by predominantly technical viewpoints, limiting holistic governance approaches. Pathways for Inclusion The AI Dialogue under the United Nations can address these gaps through: 1) Targeted capacity-building and funding support to enable participation from developing countries and under-resourced institutions 2) Regional and multilingual consultation processes to incorporate diverse linguistic and cultural perspectives 3) Structured civil society engagement mechanisms, including dedicated forums for affected communities 4) Interdisciplinary integration, ensuring that legal, ethical, and socio-economic expertise informs technical discussions 5) Digital participation platforms to broaden access beyond traditional diplomatic or institutional channels Ensuring inclusive participation is not only a matter of representation but also essential for developing legitimate, context-sensitive, and globally applicable AI governance frameworks.

To foster meaningful and dynamic engagement, the AI Dialogue should move beyond traditional plenary formats and adopt interactive, problem-solving-oriented mechanisms under the United Nations. 1. Policy Labs and Co-Creation Workshops Small, multi-stakeholder groups (governments, industry, academia, civil society) can collaboratively develop draft frameworks, model guidelines, or regulatory toolkits on specific themes such as AI safety or transparency. These labs should be outcome-oriented, with clearly defined deliverables. 2. Scenario-Based Simulations ("AI Governance Sandboxes") Participants engage in simulated real-world scenarios—such as cross-border AI incidents or algorithmic harm cases—to test governance responses. This approach helps identify regulatory gaps and promotes practical, experience-based learning. 3. Regional Dialogue Tracks Parallel regional consultations can surface context-specific challenges and priorities, ensuring that global discussions are informed by diverse socio-economic realities. Outputs from these tracks should feed directly into global deliberations. 4. Multi-Stakeholder Roundtables with Equal Voice Design Structured formats that limit dominance by any single stakeholder group (e.g., moderated equal-time interventions, curated representation quotas) can ensure balanced participation. 5. Digital Participation Platforms Interactive online platforms can enable broader engagement through: a. live consultations b. polling on policy options c. iterative feedback on draft recommendations 6. Evidence and Case Study Clinics Dedicated sessions where stakeholders present real-world use cases, regulatory experiments, or failures, enabling peer learning and cross-jurisdictional insight. 7. Iterative Drafting and Feedback Cycles Circulate draft outputs for public comment and refine them through multiple rounds, ensuring transparency and inclusivity. These formats would enable the Dialogue to transition from passive discussion to collaborative governance design and actionable outcomes.

A range of policies and practices offer concrete pathways for effective AI governance, including important contributions from India. The European Union's risk-based approach, reflected in the EU AI Act, provides a structured model for classifying AI systems by risk and imposing proportionate obligations. Similarly, the OECD AI Principles and UNESCO's Recommendation on the Ethics of AI (under the United Nations Educational, Scientific and Cultural Organization) establish widely accepted norms on transparency, accountability, and human-centric AI. India offers strong practice-oriented models. The NITI Aayog National Strategy for AI promotes "AI for All," focusing on inclusive sectoral applications. India's Digital Public Infrastructure-such as Aadhaar and Unified Payments Interface-demonstrates how interoperable, scalable platforms can enable inclusive and responsible digital innovation, with growing relevance for AI ecosystems. India has also advanced responsible AI guidance through policy discussions emphasizing safety, trust, and accountability while supporting innovation. In addition, regulatory sandboxes enable controlled testing of AI systems, balancing innovation with oversight. Open-source ecosystems and data-sharing frameworks further promote accessibility and collaborative development, particularly in resource-constrained contexts. Emerging tools such as algorithmic audits and impact assessments are strengthening accountability, while evolving data governance models (including consent-based architectures and data-sharing mechanisms) aim to balance access with privacy. Together, these approaches demonstrate that effective AI governance requires a hybrid model combining risk-based regulation, ethical standards, open infrastructures, and capacity-building adapted to national contexts while aligned with global principles.