India’s AI Summit and the theatre of power

New Delhi can convene. But can India command AI in 21st century?

New Delhi is hosting the AI Impact Summit, and the city is doing what it does best when the world arrives: assembling power, choreography, credentials, and noise. Scores of panel discussions will unfold across carpeted halls, each promising to shape the future of humanity, each producing the familiar harvest of photographs, declarations, and performative urgency. Many speakers will call it historic and transformative. Think-tanks (most of them aka new age lobbyists) will call it policy leadership. LinkedIn will call it proof of relevance.

Yet the harder question sits quietly beneath the frenzy: What, exactly, is being built here beyond the spectacle of being seen?

India’s G20 year was a triumph of diplomatic choreography, but it also revealed how easily optics substitute for outcomes. Many host cities did not emerge with meaningfully better civic life, only with vinyl coverings over what was inconvenient for the world to see, as though citizens could continue living indefinitely with the same apathetic local conditions once the delegates departed. These events do temporarily boost hospitality, and travel and tourism, but that is not the same as national transformation.

Geopolitics is not innovation. Conferences, however crowded, do not produce technological sovereignty. The risk is that we mistake the choreography of relevance for the substance of capability, and mistake the optics of leadership for the hard power of building.

There is something deeply familiar about this moment. A year ago, many of the same circuits were intoxicated by AR and VR. Before that, disinformation was the civilisation-ending obsession. Today it is AI. The topic changes, the panels remain. The same ecosystem of professional expertise rebrands itself overnight, tripping over one another to moderate, to keynote, to publish, to be quoted. The frenzy becomes self-sustaining: the event exists to justify the next event, the panel exists to validate the panellist, and the discourse exists to produce the impression that something consequential is underway.

This is not unique to AI. It is the broader pathology of global summit culture. We have watched it with climate action, with COP after COP, where declarations grow longer even as emissions grow steadier. We have watched it with G20 gatherings, where the crowd assemblage becomes the product, where the television optics matter more than the follow-through, and where everyone appears smarter in the afterglow of multilateral consensus.

Then everyone goes home. Political leaders return to domestic election cycles. Companies return to shareholder cycles. The communiqués remain framed, while reality remains unchanged.

These annual AI summits risk following the same script, except with far higher stakes, just as we have seen previous few years. Because artificial intelligence is the new infrastructure of power, it will determine which nations will emerge as AI power in years ahead.

AI is not a sector. It is the layer beneath sectors. The uncomfortable truth is that the frontier of this layer is not being built in conference halls. It is being built in a small number of Western and Chinese labs, backed by staggering amounts of capital, compute, hardware control, and audacious national-industrial intent.

The pioneers of foundational AI so far are curiously not emerging from summit podiums in the Global South. India does not lack talent. India does not lack users. India does not lack moral vocabulary. India lacks, still, the audacity of deployment at frontier scale.

We are discussing AI ethics while others are building AI supremacy. We are drafting frameworks while others are acquiring compute. We are hosting panels while others are locking down chips, cloud infrastructure, foundational models, and defence-linked applications. We risk becoming the world’s most eloquent convenor of a technology we do not command.

Of course, much of the world’s true AI expertise will never be shared on public stages, nor will frontier firms casually give away the moats that secure their dominance. That is precisely why India must begin again, and faster, building the depth and scale that makes such moats possible in the first place.

If our role is to provide the demographic scale, the data exhaust, the moral rhetoric, and the conference hospitality, while the architecture of intelligence is owned elsewhere, then the summit is a careless theatre.

There is also froth in the sector, globally and domestically. Unprofitable names have been bid up to lofty valuations on expectations that heavy spending on AI infrastructure will continue, that investment will inevitably translate into returns, that every company must now call itself an AI company. The hype cycle feeds the summit cycle. Both generate noise, both reward performance, and both risk obscuring the harder question of what is actually being built at the core.

The starting question must be brutally simple: when this summit ends, will India have moved one inch closer to real AI sovereignty? Not sovereignty as slogan, but sovereignty as capability: STEM ecosystem, Intellectual Property ownership, domestic compute, indigenous models, deep research ecosystems, hardware partnerships, serious capital pools, institutional depth, regulatory clarity that does not merely constrain but enables national advantage, and a long-term technological statecraft that outlives the next news cycle.

India represents one-sixth of humanity, much of it young. The AI century will not be kind to nations that confuse participation with influence. What will remain is the quiet, unforgiving balance sheet of technological power. India must decide whether it wishes to be a stage, or a force.

Srinath Sridharan is a corporate advisor and author of ‘Family and Dhanda’. X: @ssmumbai. Anand Venkatanarayanan is a strategic security and digital policy researcher. X: @iam_anandv.

(Disclaimer: The views expressed above are the author's own. They do not necessarily reflect the views of DH)