On 6 May 2026, UNU-CRIS participated in the 11th Multi-stakeholder Forum on Science, Technology and Innovation for the Sustainable Development Goals held in New York, New York at UNU-CPR. UNU-CRIS contributed to the session entitled, "Building Government AI Capacity: Insights from the City, National and Regional Perspectives."
Summary of the event:
Artificial intelligence (AI) is rapidly transforming public administration, economic systems, and social infrastructure worldwide. As governments expand the use of AI technologies, there is growing recognition of the need not only for responsible regulation, but also for stronger institutional and technical capacities to govern and deploy AI effectively in the public interest. At the same time, concerns over misinformation, cyber-enabled manipulation, and other forms of technological misuse have intensified, particularly in countries with limited capacities. In this context, the side event provided a timely platform to examine approaches to building government AI capacity at the city, national, and regional levels, while promoting inclusive international cooperation and addressing digital divides.
The discussion closely aligned with the theme of the 2026 STI Forum by examining how science, technology, and innovation can strengthen resilient, inclusive, and accountable public institutions for sustainable development. The event also contributed to ongoing discussions under the Global Digital Compact, including preparations for the first Global Dialogue on AI Governance and the establishment of the Independent International Scientific Panel on AI. By sharing practical experiences and policy perspectives, the event aimed to identify good practices and lessons for responsible and inclusive AI governance, particularly contributing to SDGs 9, 11, and 17.
Key issues discussed:
- Speakers emphasized that building government AI capacity requires strengthening institutional, technical, human, and governance capacities across the public sector, while integrating AI into broader digital transformation efforts with transparency, accountability, resilience, and public trust.
- Participants examined institutional models such as Chief AI Officers, dedicated AI units, coordination mechanisms, and technical expert deployment, stressing that effective AI governance depends on governments’ operational capacity to continuously manage and improve AI systems.
- At the city level, Tokyo’s experience highlighted the complexity of deploying AI across diverse settings from densely populated urban centers to remote islands with sparse populations. Participants noted that Tokyo’s lessons could inform global governance discussions and emphasized interoperable infrastructure, open-source approaches, and partnerships supporting resource-constrained municipalities.
- Discussions on the EU AI Act highlighted difficulties in aligning regulatory ambitions with implementation capacity. Participants also stressed the constructive role that regional organizations can play in supporting AI capacity-building, coordination, and knowledgesharing, particularly in developing countries prioritizing AI access and utilization.
- Participants stressed that AI-related risks, including misinformation, cyber threats, and infrastructure disruptions, as well as exclusion or marginalization of certain groups, require stronger technical expertise and proactive risk management. Speakers underscored that AI capacity-building should help narrow digital divides through inclusive and context-specific approaches addressing disparities in infrastructure, connectivity, skills, financing, and computing access.
- Human capital was identified as a major bottleneck. Speakers highlighted the need to attract and retain young talent, strengthen academia partnerships, promote flexible public-private career pathways, and create enabling public-sector environments for technical experts.
Key recommendations for action:
- Governments should adopt comprehensive, whole-of-government approaches to AI capacity-building that integrate institutional reform, technical expertise, human capital, digital infrastructure, and risk management into broader digital transformation strategies for sustainable development.
- Member States and international partners should strengthen mechanisms for sharing practical experiences, institutional models, and operational lessons on integrating AI expertise into public administration through AI offices, advisory networks, and coordination structures.
- AI security should be considered as broader digital resilience frameworks supported by international cooperation and AI safety networks, with attention paid to gender dimensions. Governments should strengthen national and cross-border AI risk management through investments in AI security, interoperable infrastructure, and coordination mechanisms addressing emerging threats.
- Greater investment is needed in public-sector AI talent development, including AI literacy for policymakers, technical training for civil servants, stronger collaboration with academia, and flexible talent exchange mechanisms. Participants also emphasized reforming rigid institutional structures to enable effective collaboration between policymakers and young technical experts.
- Participants shared the understanding that the discussions were grounded in the 2024 Global Digital Compact, reaffirmed the importance of the Independent International Scientific Panel on AI, and contributed to preparations for the first Global Dialogue on AI Governance in July 2026. They also agreed on the importance of continuing multidimensional discussions on AI capacity-building across international, regional, national, and city-level institutions.
