193 Nations, One Table: The UN's First Standing AI Governance Dialogue Opens in Geneva

A permanent seat for AI at the UN

On July 6, delegates gathered at the Palexpo convention centre in Geneva for the opening session of the Global Dialogue on AI Governance, the first meeting of what the United Nations intends to be a standing forum on artificial intelligence. Unlike the one-off summits and voluntary pledges that have defined AI diplomacy so far, this Dialogue is a permanent fixture: it was established by UN General Assembly Resolution A/RES/79/325, following the commitments made in the Global Digital Compact, and it is designed to reconvene on a regular basis. According to the ITU, a second session is already slated for New York in May 2027.

The framing, repeated across the UN's own materials, is deliberate. This is described as "the United Nations platform where all governments, private sector, academia and civil society will convene" on AI — with an emphasis on all. The stated purpose, per UNESCO, is to ensure that governance "reflects the priorities of all nations, not just the most technologically advanced." In a field where meaningful frontier development is concentrated in a handful of companies across two countries, that ambition is either the whole point or the central problem, depending on your vantage.

Who is actually in the room

The Dialogue is co-chaired by Ambassador Rein Tammsaar of Estonia and Ambassador Egriselda López of El Salvador — a pairing that signals the event is meant to be led by smaller states rather than the AI superpowers. Behind them sits a joint secretariat drawn from four UN bodies: the International Telecommunication Union (ITU), UNESCO, the UN Office for Digital and Emerging Technologies (ODET), and the Executive Office of the Secretary-General.

The guest list is high-level. UN Secretary-General António Guterres, UNESCO Director-General Khaled El-Enany, the President of the General Assembly, the ITU Secretary-General, and the UN Special Envoy for Digital Technologies are all listed among participants, alongside private-sector, academic, and civil-society representatives. The two-day event is structured around a high-level segment, thematic sessions, and side events. It is not happening in isolation, either: a single media accreditation covers three overlapping Geneva gatherings — the AI Dialogue (6–7 July), the World Summit on the Information Society Forum 2026 (6–10 July), and ITU's AI for Good Global Summit (7–10 July). For one week, Geneva is effectively the world's AI-policy capital.

The scientists' warning

The most quotable moments came not from diplomats but from scientists. The Independent International Scientific Panel on Artificial Intelligence — the UN's newly created expert body, co-chaired by deep-learning pioneer Yoshua Bengio and Nobel Peace Prize laureate and journalist Maria Ressa — is presenting its work during the Dialogue. Per UN News, Bengio delivered the sober core of the message: "science currently cannot guarantee that as capabilities continue to increase, AI will not cause catastrophic harm."

Ressa focused on a nearer-term danger — the collapse of shared reality. "If you can't tell fact from fiction, you cannot have a democracy," she warned, describing the information-integrity crisis as an "information Armageddon." Tammsaar, the Estonian co-chair, added the geopolitical dimension, noting that AI "could also be used for coercive purposes, to erode trust in governments, undermine democratic structures."

Two threads run through all of this. One is capability risk — the possibility that systems outrun our ability to control or even evaluate them. The other is the AI divide: the concern that development concentrated in the US and China leaves most of the world dependent, without the compute, data, or institutional capacity to shape the technology or capture its benefits. The Dialogue's design — small-state co-chairs, universal membership, an emphasis on equitable access — is a direct response to that second worry.

Why this is not the same as last week's panel report

It's worth being precise here, because two UN AI milestones landed within days of each other. On July 1, the Independent Scientific Panel published its first report — the assessment that generated headlines about guardrails failing to keep pace. That report is a scientific product: an attempt to give policymakers a common evidence base, roughly analogous to what the IPCC does for climate.

The Global Dialogue that opened July 6 is a different animal: it is the political venue where that evidence is meant to be turned into cooperation among states. The report informs; the Dialogue is where governments are supposed to act on it. The panel's findings are, in effect, the opening brief handed to the diplomats. Reading the two together is the point — the science says risk cannot currently be ruled out, and the Dialogue is the mechanism the UN is betting can build a shared response.

Hype versus substance

Here is where skepticism is warranted. The Dialogue is a forum, not a regulator. Nothing in the source material suggests it has binding authority, enforcement power, or a mandate to set rules — its listed functions are to convene, to "share best practices and lessons learned," and to "build common approaches." The UN has run this play before with mixed results: inclusive, consensus-driven processes are strong on legitimacy and slow on teeth. With frontier labs shipping new models on a cadence measured in weeks, a body that meets roughly once a year in different cities faces an obvious pacing problem — the very mismatch the scientific panel flagged.

The counterargument is that legitimacy is itself the scarce resource. National regulators and bilateral deals cannot, by definition, produce the universal buy-in that the AI divide demands. If any institution can credibly claim to speak for the 193 countries that are not home to a frontier lab, it is this one. Whether that voice translates into anything the labs or their host governments must heed is the open question — and one this first session, focused on convening rather than deciding, was never going to answer.

The takeaway

The Global Dialogue on AI Governance is the UN planting a permanent flag: AI is now a standing item on the international agenda, with its own forum, its own scientific panel, and its own diplomatic machinery. That is a real institutional milestone, and the scientists' warnings — that catastrophic harm cannot be ruled out, and that the information ecosystem is already fraying — give it genuine urgency. But an institution's existence is not the same as its influence. The Dialogue arrives with legitimacy and reach but, on the evidence available, without binding power or a cadence that matches the technology. The honest read: Geneva this week is a beginning worth taking seriously, and a reminder of how far the world still is from governance that can keep up. Watch what, if anything, the May 2027 New York session is empowered to actually decide.