by Beatriz Busaniche, from Ginebra.
With an attendance of approximately 4,000 participants from around the world, the first segment of the United Nations-organized Global Dialogue on AI Governance concluded today in Geneva. The event was one of three major gatherings taking place in Geneva this week, alongside the WSIS Forum 2026 and the AI for Good Global Summit, which is organized by the International Telecommunication Union (ITU).
Governments from across the globe convened during these sessions to establish a series of positions regarding the current state of AI and the future construction of appropriate governance frameworks to protect human rights. The event featured active participation from various private sector companies, as well as civil society, the technical community, and academia.
On the opening day, the “Preliminary Report of the International Scientific Independent Panel on Artificial Intelligence” was released—a collaborative study developed by 40 experts from diverse fields across the planet. This document immediately became the common baseline for discussions centered around critical areas such as regulation, public policy, investment, and development strategies. It also highlighted the need for infrastructure development, as well as the protection of minorities, children, and adolescents, among other recurring themes. The urgent need to halt the development and deployment of AI in warfare was another prominent topic among those warning of the dangers of advancing without proper oversight and regulations.
The preliminary report offers a comprehensive overview based on empirical research regarding the global state of AI. While it highlights the capabilities and advancements that can stem from its transformative potential, it rigorously addresses severe systemic risks. These include the erosion of information integrity, the creation of deepfakes that undermine public debate and the safety of specific social groups, sycophantic behavior in AI systems, irresponsible use among children and within educational systems, the concentration of power among private corporate actors, and the challenges faced by the Global South in meaningfully integrating into development and adoption processes.
Regarding this last point, numerous African delegations emphatically argued that access is not the core issue when discussing AI; rather, it is the opportunity to develop local capabilities. There were also strong voices advocating for open-source, frugal, and low-energy consumption models. The environmental impact of AI was another frequent topic across various panel discussions, as was the urgent need to safeguard the educational system to avoid repeating past mistakes. A resonant call to stop experimenting on children by introducing unauditable and uncontrollable technologies into schools was delivered by educational organizations and UNICEF’s own delegation.
While this meeting is not expected to yield any binding resolutions, it serves to establish common working ground, provide an evidence-based diagnosis, and build strategies to challenge the concentrated power of Big Tech and address the lack of state capacity to establish adequate policies for protecting human rights amid the rapid development and adoption of these technologies.
This was the first of two meetings organized by the UN. The next one will take place at the New York headquarters in May of next year, allowing time to develop further positions that will contribute to building public policies suited to the challenges of our times.
