When India's Minister of State for External Affairs Kirti Vardhan Singh stood before the High-Level Plenary of the inaugural UN Global Dialogue on AI Governance in Geneva, he delivered a message that the Googles, Metas, and OpenAIs of the world would prefer to keep at the level of polite diplomatic abstraction: the countries bearing the sharpest risks of AI-driven disruption are not the countries writing the rules.

Singh's address was precise. AI governance, he argued, must be rooted in meaningful human oversight, must uphold human rights, and must prevent misuse — not as aspirational footnotes but as structural requirements of any legitimate international framework. He called explicitly for developing nations to have a genuine, not ceremonial, role in shaping the technology's trajectory. The word that kept appearing in his remarks was "inclusive" — and in diplomatic language, when a word appears that often, it means its opposite is already happening.

The backdrop matters enormously here. The UN Secretary-General has repeatedly flagged AI as one of the defining governance challenges of this decade, and the Geneva dialogue was the first formal attempt to convene member states around a shared multilateral framework. That it happened at all is significant. That it happened without any binding mechanism, enforcement architecture, or defined accountability chain is equally significant — and far less reported.

India's position is not naïve idealism. New Delhi has its own advanced AI ambitions, a rapidly expanding domestic tech sector, and a calculated interest in ensuring that whatever global norms emerge do not simply codify the competitive advantages of the United States and China. Singh's remarks should be read in that light: this is diplomacy with a national interest attached, but the national interest happens to align with a genuine structural problem. When AI systems trained predominantly on Western and Mandarin-language data get deployed in multilingual, lower-income markets, the error rates go up, the cultural distortions compound, and the accountability chains dissolve at the border.

China, notably, was making near-identical arguments in the same room. Beijing's delegation called for bridging AI divides and explicitly rejected what it framed as pressure to "choose sides" in an emerging US-China technology bifurcation. On the surface, this produces an unusual alignment: India and China, two countries with a live border dispute and deep strategic rivalry, both pushing the same rhetorical line at a UN forum. The convergence is real but should not be mistaken for coordination. Both countries are playing separate games that happen to produce the same opening bid.

The harder question — the one that the plenary format is structurally designed to avoid — is accountability. When an AI system deployed by a private company domiciled in one jurisdiction causes documented harm in another, who answers for it? Singh's statement that governance must "prevent misuse" sounds self-evident until you ask: prevent it how, enforced by whom, with what remedy available to the person harmed? No delegation in Geneva answered that. No delegation was asked to.

What the Global Dialogue did produce is a record. Statements entered into the UN process become part of the official documentary architecture that later negotiations reference. India's call for human-centric governance and China's call for equity in AI development are now on that record. The US position — to the extent it was articulated publicly at this forum — favors industry-led governance with government guardrails, a framework that has served American platform companies extremely well in every previous technology cycle.

The gap between a UN dialogue and a UN framework with teeth is measured in years of negotiation and, more precisely, in whether the countries that control the compute infrastructure, the foundation models, and the distribution platforms decide they have more to gain from multilateral rules than from the current arrangement. History suggests they will need more pressure than a Geneva plenary to reach that conclusion. What India and the broader developing-nation bloc accomplished in Geneva is the beginning of that pressure — formal, documented, and increasingly hard to ignore.