UN Meetings Expose the Tensions That Could Shape Global AI Governance

Geneva's CSTD meetings exposed the fault lines—equity, governance philosophy, sector impacts—that could dominate July's AI governance summits.

Helena Mannion

UN meetings in April exposed four tensions shaping AI governance: who has the infrastructure and capacity to develop AI, how data should be governed across borders, whether developing countries can access and transfer technology, and whether regulation steers or constrains innovation. These questions will dominate July's global AI summits—and the answers could shape international AI policy for years.

UN meetings in Geneva from April 20–24 exposed multiple tensions—equity gaps, data governance, infrastructure concentration, capacity disparities—that could shape global AI regulation for years.

The Commission on Science and Technology for Development (CSTD) convened its 29th Session under the theme “Science, Technology and Innovation in the Age of AI.” The agenda spanned data governance frameworks, technology transfer commitments, capacity building, and the 20-year follow-up to the World Summit on the Information Society. But threading through the discussions was a fundamental question: as AI reshapes development, who gets to build and govern it—and who is excluded?

The core tensions

Four interconnected problems emerged from the CSTD discussions.

The AI adoption divide. AI adoption in the Global North is growing nearly twice as fast as in the Global South. The ITU Secretary highlighted stark disparities: only 33 countries have inference-grade AI compute capacity, only 24 have training-grade compute, and the Global South—home to half the world’s internet users—holds less than 10% of global data center capacity. A single country hosts ten times more data centers than any other; a single company manufactures nearly every leading AI chip.[1] This concentration may shape who has capacity to develop AI—and who remains dependent on others’ decisions.

Data governance and responsible AI. The meeting devoted significant attention to data as the foundation for AI. Without clear governance frameworks for data—how it flows across borders, who owns it, how it’s protected—responsible AI deployment becomes difficult. Developing countries in particular may struggle to shape these frameworks if not included in the conversation.

Technology transfer and capacity building. Ministers from developing countries consistently raised a core challenge: access to advanced technologies remains limited, and partnerships often don’t build lasting institutional capacity. Training, infrastructure, and resources needed for countries to participate meaningfully in the AI economy remain concentrated in wealthy regions.

A governance philosophy divide. The dominant metaphor shaping policy is that regulation acts as the “brakes” on innovation. At the Digital World Conference (DWC): AI for Social Development, Geoffrey Hinton challenged this sharply: regulation should work like a steering wheel—an active instrument for directing how AI develops.[2] This seemingly technical distinction may shape whether countries can steer AI toward their own priorities, or whether they simply accept trajectories set elsewhere.

Two summits, one significant moment

The CSTD meetings set the agenda for what comes next. In July, two major events converge in Geneva:

  • July 6–7: UN Global Dialogue on AI Governance — Member States, academia, civil society, and the private sector will directly confront these tensions: How do we bridge AI divides? What does trustworthy AI governance look like? How do we protect human rights in an AI-driven world?
  • July 7–10: AI for Good Global Summit returns to Geneva, open to the public and continuing the city’s role as the global focal point for AI governance conversations.

Back-to-back, these events create a focused window: the conversations that have been fragmented across policy communities, technology companies, and civil society groups could come into sharp focus in a single place.

What this means

The tensions identified at the CSTD are not abstract policy disputes. They determine who can access the infrastructure to develop AI, who influences governance frameworks, and who gets to participate in setting the rules.

For Swiss organisations and compliance teams, these conversations matter. The data governance standards, technology transfer expectations, and capacity building commitments being debated in Geneva will likely become the technical norms that reach Swiss markets first—through bankers, insurers, and tech firms embedded in global networks.

More fundamentally: the international community faces a choice about AI’s future. Can it build both the engine and the steering wheel together—innovation accelerating development, governance directing it toward shared priorities?

Sources

[1] CSTD 29th Session — Opening Plenary and Ministerial Roundtable — Remarks by Doreen Bogdan-Martin, ITU Secretary-General. April 20–24, 2026.

[2] Geoffrey Hinton on AI governance — Remarks at the Digital World Conference (DWC): AI for Social Development, co-organized by UNRISD in Geneva. UN Radio, April 23, 2026.