The United Nations has said what governments in most of the developing world already sense but rarely say aloud: artificial intelligence is moving faster than the institutions meant to govern it. The gap between the technology's trajectory and the regulatory architecture trying to contain it is not a temporary lag to be corrected with a few committees and a ministerial communiqué. It is a structural condition — and the longer it persists, the more the rules of the AI era will be written by the actors who got there first, for the benefit of the actors who got there first.
That is the real message inside the UN's alarm. The technology itself — its capacity to automate cognition, compress decision cycles, and operate across jurisdictions without friction — is only half the story. The other half is geopolitical: whoever codifies AI governance at the multilateral level is effectively setting the terms of participation for every nation that builds, deploys, or regulates AI afterward. Right now, the contenders for that standard-setting role are the European Union, with its risk-classification architecture already in force, and the United States, which prefers voluntary frameworks and bilateral agreements. Neither model was designed with Bengaluru or Bhopal in mind.
The Westphalian Problem in Silicon
There is a structural irony worth pausing on. The international system was built, after centuries of war, around the principle that states are sovereign over their own territory — that external powers cannot impose legal obligations without consent. AI, as a technology, respects none of those boundaries. A model trained in San Francisco, hosted on servers in Ireland, and fine-tuned with data scraped from a dozen countries operates in every jurisdiction simultaneously, before a single regulator has had the chance to examine it. The Westphalian premise — that a state can govern what happens within its borders — does not transfer cleanly to a technology whose outputs arrive before its governance does.
This is why the UN's call for urgency is not merely procedural. It reflects a recognition that the window for genuinely multilateral AI governance — governance that reflects the interests of 193 member states rather than the G7's regulatory preferences — is closing. The EU's AI Act is already law. US executive orders on AI have already shaped how the largest AI companies operate. If the UN process does not crystallise into something with teeth before those bilateral and regional frameworks become the de facto global standard, the rest of the world will simply be harmonising with rules it did not write.
India's Scale, India's Stake
The country that has the most to gain from getting this right — and the most to lose from getting it wrong — is India. The IndiaAI Mission, as the UN discussion on AI governance intensifies, carries ₹10,371 crore in government funding, a stated ambition to build indigenous compute capacity, and a framework that explicitly resists the EU's risk-based compliance model in favour of an approach centred on capacity-building and responsible deployment. That is not a small distinction. The EU's model classifies AI applications by risk level and attaches compliance obligations accordingly — an architecture that suits large, well-capitalised firms with dedicated legal teams. For India's startup ecosystem and its tens of thousands of small and medium enterprises beginning to integrate AI into manufacturing, agriculture, and services, that compliance overhead could be genuinely prohibitive.
The IT services export sector sits at the sharpest edge of this tension. Generative AI is compressing the kind of business-process work — customer support, back-office automation, code generation — that has underwritten the growth of India's tech industry for three decades. The disruption is structural, not cyclical. No governance framework will reverse it. But governance frameworks will determine whether Indian companies can compete in the AI-augmented version of that market on equal terms, or whether they find themselves locked out of key jurisdictions by compliance regimes calibrated to the operational realities of firms headquartered in the OECD.
The Rule-Taker Trap
Arvind Gupta, writing through the Observer Research Foundation, has argued with considerable force that India must treat AI governance as a norm-shaping exercise rather than a norm-adoption exercise — that the alternative, becoming a rule-taker in a domain this consequential, amounts to a new form of technology dependency dressed in the language of interoperability. The analogy to earlier technology governance debates is uncomfortable but precise. In the 1990s and early 2000s, India engaged with intellectual property and internet governance frameworks largely reactively, discovering the costs of externally set standards after they were already embedded in trade architecture. The TRIPS Agreement's pharmaceutical provisions — and the decade of litigation that followed — are the canonical example of what happens when a large, capable country lets the drafting happen without it.
AI governance is moving through a similar inflection point, and the drafting window is still, barely, open. India co-sponsored the March 2024 US-led General Assembly resolution on safe, secure, and trustworthy AI — a constructive gesture, but a supporting role. The stronger play, and the one that India's scale and democratic credibility make genuinely available, is to table a concrete counter-proposal: one centred on open-source access, interoperability, and technology transfer for developing nations rather than liability frameworks designed around the risk profiles of large Western models. India's G20 Presidency produced a set of AI governance principles oriented around inclusion and access — the analytical infrastructure for such a proposal already exists. The question is whether those principles get translated into negotiating text before someone else's negotiating text becomes the baseline.
The Coalition India Could Lead
No single country dominates multilateral norm-setting through unilateral assertion; it happens through coalitions. India's natural interlocutors for an alternative AI governance framework are not the obvious ones. Japan and Singapore bring technical credibility and democratic legitimacy. The UAE has moved with unusual speed on both AI investment and governance infrastructure. Brazil carries Global South weight in multilateral forums comparable to India's own. A coalition built around those five countries — with India setting the agenda — would represent a significant share of the world's AI talent, AI investment, and AI-using population. It would not be a bloc in the Cold War sense; it would be a standard-setting coalition, the kind that shapes international agreements before they are finalised rather than lobbying for exemptions after the fact.
The structural fault-line in India's current posture is that its influence has been exercised reactively — co-sponsoring, endorsing, cautioning — rather than proactively. Nasscom and the Confederation of Indian Industry have formally cautioned against importing the EU AI Act's compliance architecture. Parliamentarians have called for equitable UN representation in AI governance bodies. The Ministry of Electronics and Information Technology has built a framework that deliberately diverges from the Western regulatory mainstream. The analytical consensus, in other words, points in one direction. What it awaits is the diplomatic initiative to match it.
The UN's warning that governments cannot keep pace with AI is accurate, but it is also, in a precise sense, a political opportunity framed as a crisis. The rules of the AI era have not yet been written. The institutions meant to write them are still under construction. India enters this moment with more cards than it typically plays — a credible digital infrastructure, a democratic mandate, a massive domestic market, and a Global South standing that no Western power can replicate. Whether that translates into genuine agenda-setting at the UN, or into another round of carefully worded co-sponsorships, will determine whether India's AI ambitions are built on a foundation it helped architect or one it inherited from elsewhere.
