Strengthening Global AI Governance: Insights from the First UN Global Dialogue on AI Governance and the AI for Good Global Summit

IRIS Sustainable Development participated in the Global Dialogue on AI Governance and the AI for Good Global Summit in Geneva, engaging alongside governments, international organizations, the private sector, academia, and civil society to advance discussions on responsible, inclusive, and evidence-based AI governance.

The Global Dialogue on AI Governance is a United Nations platform established by the UN General Assembly through Resolution A/RES/79/325 and rooted in the Global Digital Compact. It provides an inclusive forum where all 193 UN Member States, together with technology companies, civil society, academia, and the technical community, convene to strengthen international cooperation and coordinate approaches to AI governance. As the first Global Dialogue on AI Governance, it marked an important milestone in institutionalizing multilateral dialogue on artificial intelligence within the United Nations system and reaffirmed the need for a governance architecture capable of responding to the global nature of AI.

Held alongside the Dialogue, the AI for Good Global Summit, organized by the International Telecommunication Union (ITU) in partnership with United Nations agencies, has become the leading global platform for demonstrating how artificial intelligence can accelerate progress towards the Sustainable Development Goals (SDGs). Bringing together policymakers, researchers, innovators, entrepreneurs, industry leaders, and international organizations, the Summit showcased AI applications addressing challenges in healthcare, education, climate action, disaster risk reduction, digital inclusion, agriculture, and sustainable economic development. The Summit highlighted that effective AI governance and responsible innovation must evolve together if AI is to deliver meaningful societal benefits.

One of the defining themes of the Dialogue was interoperability. Participants emphasized that building safe, secure, and trustworthy AI does not require countries to adopt a single global regulatory model. Rather, it requires governance frameworks capable of operating coherently across jurisdictions while respecting different legal traditions, institutional capacities, policy priorities, and societal contexts. The objective is therefore not regulatory uniformity, but practical cooperation.

Achieving interoperability requires more than high-level principles. It depends on developing shared terminology, comparable approaches to risk classification, interoperable incident-reporting mechanisms, common technical standards, stronger scientific evidence, and independent evaluation methodologies. Such mechanisms can reduce regulatory fragmentation, strengthen accountability, and facilitate international cooperation while preserving national sovereignty and regulatory autonomy. As AI systems increasingly transcend national borders, interoperability is becoming an essential foundation for effective global AI governance.

The discussions also underscored the importance of a more equitable and inclusive governance architecture. As the development of frontier AI remains concentrated within a relatively small number of countries and companies, participants stressed that international governance should not simply reproduce existing technological asymmetries. Instead, developing countries should become co-creators of global AI governance through sustained capacity development, technology cooperation, multilingual evaluation methodologies, knowledge exchange, and meaningful participation in international standard-setting processes. Inclusive governance is essential not only to bridge the global AI divide, but also to ensure that governance frameworks reflect diverse societal priorities, cultures, and development needs.

Beyond governance structures, the Dialogue highlighted that AI policy must remain firmly grounded in the public interest. While artificial intelligence presents unprecedented opportunities to advance scientific discovery, healthcare, education, climate resilience, disaster preparedness, and sustainable development, its rapid evolution also introduces significant societal challenges. Participants reflected on AI's implications for democratic institutions, information ecosystems, labour markets, critical infrastructure, and cultural and creative industries, emphasizing that innovation must be accompanied by robust safeguards, transparency, accountability, and respect for fundamental rights.

Particular attention was given to children's rights in the digital age. AI has the potential to transform education, improve learning outcomes, and expand access to knowledge. At the same time, it can expose children to manipulation, exploitation, harmful content, algorithmic bias, and risks to their mental, cognitive, and social development. Ensuring that AI systems are designed, deployed, and governed in ways that protect children's rights, safety, well-being, and development emerged as a shared international priority. Protecting vulnerable populations must therefore become an integral component of responsible AI governance rather than an afterthought.

The discussions also explored the increasingly complex relationship between AI and societal resilience. AI can strengthen the operation of critical infrastructure, improve climate modelling, optimize energy systems, and enhance disaster response. Yet the same technologies can also be weaponized to identify vulnerabilities in digital and physical infrastructure, amplify cyber threats, increase pressure on natural resources through growing computational demands, and disrupt information ecosystems. Similarly, while generative AI is built upon vast repositories of human creativity and journalistic content, it is also transforming cultural production and raising important questions regarding intellectual property, fair remuneration, and the long-term sustainability of creative industries.

At IRIS Sustainable Development, one of the key messages we emphasized is that cybersecurity in the age of AI is no longer only about protecting networks and systems. It is equally about safeguarding the integrity of the information ecosystem.

As generative AI dramatically lowers the cost of producing highly convincing synthetic content at scale, the principal challenge extends beyond distinguishing truth from falsehood. It lies in preserving trust in the information upon which individuals, institutions, and governments rely to make informed decisions, participate in democratic processes, and respond effectively during periods of crisis. Information integrity is therefore becoming an increasingly important dimension of both cybersecurity and democratic resilience.

Strengthening societal resilience requires investment across the entire digital ecosystem: secure digital infrastructure, trusted communication channels, robust data governance, independent evaluation and assurance mechanisms, transparent AI development, and local capacities to understand, assess, and deploy AI responsibly. Governance frameworks must evolve alongside technological innovation to ensure that trust remains embedded within digital societies.

The discussions in Geneva reaffirmed that effective AI governance cannot be achieved through technological innovation or regulatory action alone. It requires sustained multilateral cooperation, evidence-informed policymaking, interoperable governance frameworks, and continuous engagement among governments, international organizations, industry, academia, and civil society. As AI systems become increasingly capable and globally interconnected, ensuring that governance remains inclusive, transparent, grounded in internationally shared values, and responsive to societal needs will be fundamental to maximizing the benefits of AI while mitigating its risks.

Beyond the specific policy discussions, the Dialogue highlighted a broader reality: AI governance is no longer a future challenge—it is a present responsibility. The decisions taken today will shape not only the trajectory of AI technologies but also the future of trust, security, democratic governance, economic opportunity, and sustainable development. The success of international AI governance will therefore depend not on identical regulations, but on the ability of diverse governance systems to communicate, exchange evidence, recognize one another's safeguards, and cooperate in addressing shared challenges.

For IRIS Sustainable Development, participating in these discussions reinforced the importance of connecting research, policy, and implementation. Responsible AI governance cannot be achieved by any single stakeholder or nation. It requires a shared commitment to multilateral cooperation, scientific evidence, human rights, and inclusive development. As AI continues to transform societies, we remain committed to contributing to international efforts that ensure AI is governed in ways that strengthen resilience, protect fundamental rights, and advance the public interest while supporting the achievement of the Sustainable Development Goals.