Intuitive.ai / GovStack Women in GovTech Challenge / UN CSTD Multi-Stakeholder Working Group on Data Governance

A successful first Global Dialogue on AI Governance should move beyond broad principles and produce concrete areas for international cooperation. The Dialogue should establish a shared understanding that trustworthy AI depends not only on models and regulation, but also on the quality, governance, interoperability, and ownership of data. Success would include: Agreement on common principles for data governance, interoperability, transparency, and accountability in AI systems Stronger cooperation between governments, technical experts, civil society, academia, and the private sector Practical recommendations for supporting developing countries through capacity building, open standards, and digital public infrastructure Recognition of the importance of digital sovereignty and the need to avoid new forms of technological dependence Greater inclusion of women and underrepresented groups in AI governance discussions The Dialogue should also create a pathway for continued collaboration beyond the first meeting, including working groups, technical exchanges, and implementation-oriented follow-up.

• Safe, secure and trustworthy AI
• Interoperability of governance approaches
• Transparency, accountability, and human oversight
• Open-source software, open data and open AI models

I selected these priorities because they are foundational to building AI systems that are effective, trustworthy, and aligned with the public interest. Safe, secure, and trustworthy AI is essential because weak data governance and poor-quality systems can create risks including bias, exclusion, misinformation, and security vulnerabilities. Interoperability of governance approaches is increasingly important because AI systems operate across borders, sectors, and platforms. Countries need common standards and interoperable frameworks to enable cooperation while reducing fragmentation and duplication. Transparency, accountability, and human oversight are critical to ensuring that AI systems remain understandable, contestable, and subject to meaningful public control. Finally, open-source software, open data, and open AI models can support innovation, transparency, and digital sovereignty, particularly for developing countries and public-sector institutions.

One important cross-cutting issue that is not sufficiently reflected is data governance and digital sovereignty. AI governance discussions often focus on models and regulation, while overlooking the data foundations that determine whether AI systems are trustworthy and equitable. Questions of who owns data, how it is governed, how it moves across systems, and whether countries retain meaningful control over their digital infrastructure are becoming increasingly important. Another emerging issue is interoperability between digital public infrastructure and AI systems. As governments adopt AI in areas such as health, education, social protection, and digital identity, there is a growing need for open standards, interoperable systems, and public-interest digital infrastructure. There is also a need for greater participation of women, underrepresented communities, and technical practitioners from the Global South in shaping AI governance. Without this, there is a risk that AI governance will reflect only the perspectives of a small number of countries and companies.

In the technology and public-sector ecosystem, one of the most significant governance gaps is the lack of common standards for data governance, interoperability, and accountability in AI systems. Many organizations and governments are rapidly adopting AI, but often without clear frameworks for data quality, transparency, human oversight, or secure data sharing. In India and across many developing countries, this creates both risks and opportunities. Public systems increasingly rely on AI for areas such as healthcare, social protection, education, agriculture, and digital identity. However, data often remains fragmented across institutions and platforms, making it difficult to build reliable, interoperable, and trustworthy systems. The absence of interoperable governance approaches also creates a risk of dependence on a small number of global technology providers and closed systems. This can limit digital sovereignty and reduce the ability of countries to shape AI in ways that reflect local needs, languages, and priorities. At the same time, there is a major opportunity to build more inclusive and resilient systems through open standards, digital public infrastructure, and open-source technologies. India's experience with digital public infrastructure demonstrates that interoperable and scalable public systems can support innovation while maintaining public trust. There is also a significant opportunity to involve more women, technical experts, and underrepresented groups in AI governance. Greater diversity in the design of AI systems and governance frameworks can help reduce bias and ensure that AI serves broader social and economic goals.

The AI Dialogue can play an important role by creating a neutral and inclusive space where governments, technical experts, civil society, academia, and the private sector can work toward common approaches to AI governance. International cooperation is increasingly necessary because AI systems operate across borders, sectors, and platforms. No single country or institution can address issues such as data governance, interoperability, transparency, digital sovereignty, and AI safety alone. The Dialogue can help advance cooperation by: Identifying common principles and practical areas for alignment Encouraging interoperability between different governance approaches Sharing best practices and lessons learned across regions Supporting capacity building, particularly for developing countries Creating opportunities for technical collaboration on standards, open-source technologies, and digital public infrastructure The Dialogue should also ensure that technical practitioners, women, and voices from developing countries are actively included. A successful Dialogue should not end with discussion alone. It should lead to ongoing collaboration through working groups, technical exchanges, pilot initiatives, and implementation-oriented follow-up.

The AI Dialogue should build upon and connect with existing international efforts rather than create a separate process. Important initiatives include: The UN CSTD Multi-Stakeholder Working Group on Data Governance at All Levels The Global Digital Compact and the Global Digital Cooperation process ITU, UNESCO, and OECD work on AI principles and governance GovStack and broader digital public infrastructure initiatives Open-source and open standards communities Regional and national AI strategies, particularly from developing countries These initiatives already provide valuable work on topics such as data governance, interoperability, digital public infrastructure, capacity building, and responsible AI. However, they often operate in separate communities or focus on individual aspects of AI governance. The added value of the AI Dialogue would be to connect these efforts, encourage greater coordination, and create a common space for practical cooperation. It can help bridge the gap between policy and technical implementation, bring together different stakeholder groups, and ensure that discussions on AI governance are more inclusive, interoperable, and globally representative.

Different stakeholders should contribute according to their expertise and experience. Governments can share policy priorities and regulatory approaches. Technical experts can provide practical insights on how AI systems work in real-world settings. Civil society can raise issues of rights, inclusion, and public trust. Academia can contribute research and evidence, while the private sector can share lessons from implementation and innovation. To make the AI Dialogue effective, the format should combine: High-level plenary discussions Smaller thematic working groups Interactive technical sessions Opportunities for written inputs and open consultation before and after the Dialogue Thematic working groups could focus on areas such as data governance, interoperability, transparency, open-source technologies, and capacity building. The Dialogue should also include structured spaces where different stakeholder groups can engage directly with one another rather than speaking separately. The process should not end after the event. A successful structure would include follow-up working groups, shared outputs, and a mechanism for continued collaboration and implementation.

Several important perspectives remain underrepresented in global discussions on AI governance. These include: Technical practitioners such as data engineers, open-source developers, and digital public infrastructure experts Women and underrepresented groups in technology Voices from developing countries and the Global South Small and medium enterprises, startups, and local innovation ecosystems Communities directly affected by AI systems, including people in rural and low-resource settings Too often, global AI discussions are dominated by a small number of governments, large technology companies, and institutions from a few regions. As a result, important concerns such as local languages, digital sovereignty, access, and the realities of implementation may be overlooked. These groups can be better included through targeted outreach, financial support for participation, multilingual engagement, open calls for written inputs, and dedicated sessions for underrepresented communities. The Dialogue should also ensure gender balance and regional diversity in speakers, moderators, and working groups.

Meaningful engagement is more likely when participants can actively contribute rather than only listen to formal statements. Useful formats could include: Small multi-stakeholder roundtables focused on specific challenges Interactive workshops and scenario exercises Technical demonstrations of open-source tools, digital public infrastructure, and AI governance approaches Case-study sessions where countries or organizations share practical lessons Collaborative drafting sessions to develop common principles or recommendations Another effective format would be "problem-solving labs," where participants from different sectors work together on a shared issue such as AI in public services, data interoperability, or digital sovereignty. Hybrid and digital participation options are also important to ensure that experts from different regions can participate meaningfully. The Dialogue could also use open online platforms before and after the event to collect ideas, enable continued discussion, and support collaboration beyond the two-day meeting.

Several existing policies and approaches provide useful foundations for effective AI governance. India's digital public infrastructure model demonstrates how open, interoperable platforms can support innovation while maintaining public trust. Systems such as digital identity, payments, and data-sharing frameworks show the value of open standards, interoperability, and public-interest digital infrastructure. GovStack is another important example. It promotes interoperable and reusable digital building blocks that governments can adapt to local needs. This approach reduces duplication, supports digital sovereignty, and helps countries build secure and resilient public systems. The UNESCO Recommendation on the Ethics of Artificial Intelligence provides an important global framework for transparency, accountability, human rights, and inclusion. Similarly, the OECD AI Principles offer guidance on trustworthy and human-centered AI. Open-source software and open standards are also critical tools for effective AI governance. Open-source approaches can improve transparency, allow independent review, reduce dependence on closed systems, and support local innovation ecosystems. Open data and interoperable frameworks are particularly valuable in areas such as healthcare, education, agriculture, and disaster response. Data governance practices are equally important. Effective approaches include: Clear rules for data ownership and access Data quality standards Privacy-preserving data sharing Transparent data lineage and provenance Human oversight and accountability mechanisms Finally, multi-stakeholder approaches such as the UN CSTD Multi-Stakeholder Working Group on Data Governance at All Levels demonstrate the importance of bringing together governments, technical experts, civil society, academia, and the private sector to develop practical and inclusive governance solutions.